Russian State Geology Prospecting University
COMPLEX OF
SPECTRAL-CORRELATION ANALYSIS
OF GEOPHYSICAL EVIDENCE
«COSCAD 3Dt»
USER’S MANUAL
PART I
Version 2007.1
Coscad 3Dt
2
1. Computer technology COSCAD 3Dt. ............................................................................................................................................. 4
1.1. Program overview .................................................................................................................................................................... 4
1.2. About Authors .......................................................................................................................................................................... 8
1.3. How to install "Coscad 3Dt"? .................................................................................................................................................. 9
1.4. FAQ and Error's handling ...................................................................................................................................................... 10
2. SERVICE ...................................................................................................................................................................................... 13
2.1. Data input... ............................................................................................................................................................................ 13
2.1.1. Inputting data from “COSCAD 3Dt” - formats ............................................................................................................... 13
2.1.2. Inputting data from SURFER format .............................................................................................................................. 15
2.1.3. Inputting data from GEOSOFT format ........................................................................................................................... 15
2.1.4. Inputting data from SEGY format ................................................................................................................................... 15
2.1.5. Inputting data from INTEGRO format ............................................................................................................................ 15
2.1.6. Inputting data from RADAR format ............................................................................................................................... 15
2.2. Data output... .......................................................................................................................................................................... 16
2.2.1. Outputting data into COSCAD 3Dt format ..................................................................................................................... 16
2.2.2. Outputting data into SURFER format ............................................................................................................................. 17
2.2.3. Outputting data into GEOSOFT format .......................................................................................................................... 18
2.2.4. Outputting data into SEGY format .................................................................................................................................. 18
2.2.5. Outputting data into INTEGRO format ........................................................................................................................... 18
2.2.6. Outputting data into GRAPHER format .......................................................................................................................... 18
2.2.7. Outputting data as a common text table X,Y,Z,F (for EXCEL) ...................................................................................... 19
2.3. Irregular grids input... ............................................................................................................................................................. 19
2.3.1. Inputting of 2D irregular grids ........................................................................................................................................ 19
2.3.2. Inputting of 2D irregular grids (2D-spline) ..................................................................................................................... 20
2.3.3. Inputting of 3D irregular grids ........................................................................................................................................ 21
2.4. Interpolation of grids... ........................................................................................................................................................... 22
2.4.1. Liner interpolation ........................................................................................................................................................... 22
2.4.2. Profile spline interpolation .............................................................................................................................................. 22
2.4.3. 2D Spline interpolation ................................................................................................................................................... 22
2.4.4. Profile Fourier interpolation ............................................................................................................................................ 23
2.5. Extrapolation of grids ............................................................................................................................................................. 23
2.6. Filling of absent data .............................................................................................................................................................. 23
2.7. Fragmentation of grids ........................................................................................................................................................... 24
2.8. Inserting of grids .................................................................................................................................................................... 24
2.9. Rotating of grids ..................................................................................................................................................................... 24
2.10. Uniting of grids .................................................................................................................................................................... 25
2.11. Some transformations of data ............................................................................................................................................... 25
2.11.1. Transformations with one sign ...................................................................................................................................... 25
2.11.2. Transformations with two signs .................................................................................................................................... 26
2.11.3. Imposition of dummy-code ........................................................................................................................................... 26
2.12. Gluing of grids ..................................................................................................................................................................... 27
2.13. Printing information into file ................................................................................................................................................ 27
2.14. Adaptive equalization of seismic traces ............................................................................................................................... 27
3. VISUALIZATION ........................................................................................................................................................................ 29
3.1. Raster map .............................................................................................................................................................................. 29
3.2. Plotting of classification ......................................................................................................................................................... 30
3.3. Graphics ................................................................................................................................................................................. 31
3.4. Inverse problem ...................................................................................................................................................................... 32
3.5. Viewing in projections X, Y, Z .............................................................................................................................................. 32
3.6. Classification viewing in projections X, Y, Z ........................................................................................................................ 33
3.7. 3D Surface view ..................................................................................................................................................................... 34
3.8. Graphic’s map ........................................................................................................................................................................ 35
4. STATISTICS ................................................................................................................................................................................ 36
4.1. Statistical characteristics... ..................................................................................................................................................... 36
4.1.1. Statistical characteristics of field fragments .................................................................................................................... 36
4.1.2. Statistical characteristics in the slithering window .......................................................................................................... 37
4.1.3. Statistical characteristics in the one-dimensional dynamic window ............................................................................... 38
4.1.4. Statistical characteristics in the two-dimensional dynamic window ............................................................................... 38
4.1.5. Statistical characteristics in the window of “alive form” ................................................................................................ 39
4.1.6. Estimation of the factor dispersion in the slithering window .......................................................................................... 39
4.2. Correlation characteristics... ................................................................................................................................................... 40
Copyright © 2003 MSGPU
Coscad 3Dt
3
4.2.1. Autocorrelation function ................................................................................................................................................. 40
4.2.2. Cross-correlation function between profiles.................................................................................................................... 41
4.2.3. Cross-correlation function between fields ....................................................................................................................... 41
4.2.4. Two-dimensional autocorrelation function ..................................................................................................................... 42
4.2.5. Two-dimensional cross-correlation function ................................................................................................................... 42
4.2.6. Three-dimensional autocorrelation function ................................................................................................................... 42
4.2.7. Correlation coefficient in the slithering window ............................................................................................................. 43
4.3. Spectral characteristics... ........................................................................................................................................................ 44
4.3.1. Univariate spectrum ........................................................................................................................................................ 44
4.3.2. Two-dimension spectrum ................................................................................................................................................ 44
4.4. Gradient characteristics .......................................................................................................................................................... 45
4.5. Sounding... ............................................................................................................................................................................. 45
4.5.1. Statistical sounding ......................................................................................................................................................... 46
4.5.2. Correlation sounding ....................................................................................................................................................... 46
4.5.3. Cross-correlation sounding.............................................................................................................................................. 47
4.5.4. Gradient sounding ........................................................................................................................................................... 48
4.6. Estimation of anomalous object’s parameters. ....................................................................................................................... 48
4.6.1. Estimation by I.I. Priezzhev ............................................................................................................................................ 48
4.6.2. Estimations by A.V. Petrov ............................................................................................................................................. 49
4.6.3. Tracing of axes of anomaly ............................................................................................................................................. 49
4.6.4. Depth estimation of the main anomalous surfaces .......................................................................................................... 50
5. FILTERING .................................................................................................................................................................................. 51
5.1. One-dimension filtering... ...................................................................................................................................................... 51
5.1.1. One-dimension filtering in the fixed window.................................................................................................................. 51
5.1.2. One-dimensional polynomial filtering ............................................................................................................................ 52
5.1.3. One-dimensional adaptive filtration ................................................................................................................................ 52
5.2. Two-dimension filtering... ...................................................................................................................................................... 53
5.2.1. Two-dimension filtering in the fixed window ................................................................................................................. 53
5.2.2. Two-dimensional polynomial filtering ............................................................................................................................ 53
5.2.3. Two-dimension adaptive filtration .................................................................................................................................. 54
5.2.4. Two-dimension filtering in the window of “alive form” ................................................................................................. 55
5.3. Three-dimensional filtering... ................................................................................................................................................. 55
5.3.1. Three-dimension entropy filtering ................................................................................................................................... 55
5.4. Decomposition of fields ......................................................................................................................................................... 55
5.5. Compensating filtering ........................................................................................................................................................... 56
5.6. Calculation of instantaneous power ....................................................................................................................................... 56
6. DETECTION ................................................................................................................................................................................ 57
6.1. Detection of weak linear anomalies... .................................................................................................................................... 57
6.1.1. Method of interprofiles correlation ................................................................................................................................. 57
6.1.2. Method of self-tuning filtering ........................................................................................................................................ 58
6.1.3. Detection of multi-signs linear anomalies ....................................................................................................................... 58
6.2. Detection of weak free form anomalies... ............................................................................................................................... 59
6.2.1. Method of inverse probability ......................................................................................................................................... 59
6.2.2. Automatic variant of the method of inverse probability .................................................................................................. 60
7. COMPLEX.................................................................................................................................................................................... 61
7.1. Classification... ....................................................................................................................................................................... 61
7.1.1. Method of the common distance ..................................................................................................................................... 61
7.1.2. Method of the dynamic thickenings (К-mean-value method) ......................................................................................... 62
7.1.3. Classification by A.V. Petrov .......................................................................................................................................... 62
7.1.4. Sign classification ........................................................................................................................................................... 63
7.2. Detection of multi-signs anomalies ........................................................................................................................................ 63
7.3. Components data analysis ...................................................................................................................................................... 64
Copyright © 2003 MSGPU
Coscad 3Dt
4
1. Computer technology COSCAD 3Dt.
1.1. Program overview
COMPLEX OF SPECTRAL-CORRELATION ANALYSIS OF THREE-DIMENSIONAL
GEOPHYSICAL EVIDENCE «COSCAD 3Dt».
The efficiency of the modern geological prospecting production is defined by the introduction degree to the process
of the manufacturing and interpretation of geology-geophisical information of new computer technologies. At the
same time the specific of geological branch predestines the usage of the most varied technologies covering
practically all main trends of the modern computer industry. The enormous amounts and variety of space-indexed
geology-geophisical information requires introducing of the most modern computer systems for its collection,
keeping, systematization and organization of the efficient searching. Geographical data binding expects the usage
of geographical searching and cartographic computer systems as well as visualization devices spatially coordinated
digital information. The modern space, air, overland and deep geophysical investigations including measurements,
processing and interpretation of the most varied geophisical fields is impossible without using of special computer
technologies. Here with on a measurement stage the adequate software for the measuring equipment is necessary.
Each geophisical method requires the own software for undertaking the qualitative processing of the primary
observations. Besides the range of problems solved by means of computer technologies in the step of interpretation
are very broad. Amongst them it is possible to select two main blocks. The former includes software directed to
solve the inverse problems using the classical deterministic approach. The latter is based on a method of
probabilistic-statistical approach and the other sections of modern applied mathematics. The original usage of these
methods allow to get additional information for solving of the primary interpretation task - construction of
qualitative and greatly explicated geology-geophisical model of the concrete geological object. And it is possible to
continue the unlimited list of specific problems solved in the process of geological production on the base of the
modern computer technologies.
* * *
Regrettably, recently the share of Russian computer developments in the field of geological studies is vastly
decreased. First of all, this pertains to areas where traditionally the lag is existed - in the development of the
geographical information systems and global databases, cartographic systems and visualization facilities. On the
other hand, the colossal amount of the domestic theoretical developments in different areas of the interpretation
gatherer for previous ten years are partly realized in the modern computer technologies not yielding to west
analogues by quality and contents. Herewith recently it is exist the increase of amount developments solving the
determined range of problems of data interpretation of gravel-magnetometry, electrometry, electromagnetic
observations, processing of nucleus- radiometric data, where the Russian school is traditionally occupied the
leading positions.
One of such developments is a computer technology of the statistical and spectral-correlation analysis of
geophysical evidence "COSCAD 3Dt", intended for analysis of three-dimensional digital geographic information
by probabilistic-statistical approaches. Works of G.A.Tarhov, A.A.Nikitin, V.I.Aronov, S.A.Serkerov,
D.A.Rodionov, G.V.Demura, O.A.Demidovich and others lie at the heart of the functional filling of this
technology. For the first time a spectrum of the original geological problems, solved by means of probabilisticstatistical methods was marked exactly in these works. The analysis of methods and results (got by means of their
first programme realisation) described in these works has allowed to work out the efficient scheme of the
processing of geological-geophysical observations by probabilistic-statistical approach. In the middle eightieth the
improvement and working out of algorithms of the geophisical information processing had brought to the unique
computer technology "COSCAD 3Dt" and occupied determined place in the crude structure of probabilisticstatistical approach to interpretation of geophisical information.
First versions of "COSCAD 3Dt" were intended just for analysis of potential geophisical fields and processing of
ore-geophisical methods. At the same time a part of algorithms contained in the complex and their theoretical
motivation were adopted from seismics with conversion for correct usage in data processing of the ore-geophysics.
At present time the inverse process is occurred. The programme complex "COSCAD 3Dt" finds more and more
usage at analysis of 2D-3D seismic information by original algorithms of statistical analysis, finding of weak
anomalies, artificial perceptions and classifications. Using instruments of the weak signal finding, created for
Copyright © 2003 MSGPU
Coscad 3Dt
5
analysis of potential fields, adaptive filters, analysis and calculation of the statistic performances in sliding-scale
windows, both on the whole seismic profile and on the separate horizon are of interest of seismic parameter fields
analysis.
The programme complex of spectral-correlation data analysis "COSCAD 3Dt" is intended for processing on digital
three-dimensional regular networks of geophisical information by methods of statistical, spectral and correlation
analysis, linear optimum filtering, weak anomaly finding, classifications and artificial perceptions. "COSCAD 3Dt"
enables to conduct full spectral-correlation and statistical analysis of given geophisical data, calculate the Fourier
spectrum, correlation functions and gradient characteristics of the geophisical data.
 There is wide range of linear optimum filters in this complex. Using them you can present the source field by
set forming with consequent reduction of the share of the low frequencies.
 Unique algorithms of adaptive linear filtering are intended for processing of non-stationary fields.
 A programme realisation of the methods of cross-profile correlation, self-adjusting filtering, inverse probability
and their multivariate analogue allows solving the problems of the weak signal finding on the background of
the commensurable hindrances on amplitude.
 The algorithms of the complex analysis of several geology-geophisical signs and their derived on basis of the
recognition and classification of complex geophisical, geochemical and geological observations methods are
intended for decision of the geological zoning and mapping problems.
 The original database of the complex allows effectively working with digital space-portioned information,
organized in three-dimensional regular grids. Service functions of the database provide information exchange
between different processing modules, carry out input/output operations, allow to fragment, unite and
complement of grids, fill missing values, realize the different algebraic transformations on signs etc.
 The graphic interface of the complex "COSCAD 3Dt" is intended for screen visualisation of information from
the database in the manner of separate graphs, graphic sets, raster maps, three-dimensional objects etc.
MAIN STRUCTURED AND PITHY COMPONENTS OF THE COMPLEX
This software uses the original database intended for storage of the 3D (multilayer and multilevel) geologygeophisical information. For all functional procedures of the complex the input and output digital information is
organized as three-dimensional regular grids. Each three-dimensional grid in the database is characterized by
following parameters (shown on Fig.1.1):
 An amount of pickets on profile N;
 An amount of profiles M and number of layers NS;
 Distance between pickets (picket’s step) Dx, profiles (profile’s step) Dy and layers (layer’s step) Dz;
 Number of digital signs Kp (this parameter is use like fourth discrete dimension);
 The coordinates of bottom-left corner of the 3D grid X0,Y0,Z0.
Except this for grid orientation in space it is possible to assign two angles U1 and U2.
Accepted definition:
Grid is a 3D-cube of digital regular information (possibly has several signs).
Layer is identified the horizontal-plane cut of the three-dimensional grid.
Vertical cut is a cut oriented in planes of the same named profiles.
The parallelepiped-shaped cut in grid with size smaller then the size of full grid is called the grid fragment.
Layers can be presented as data measured on different height or depth.
A number of pickets, profiles and layers in grid are limited just only by resources of concrete computer.
On Fig.1.1 is shown direction of right-handed coordinate system (accepted in software complex) and numbering
order of pickets, profiles and layers.
Copyright © 2003 MSGPU
Coscad 3Dt
6
Z
Y
1
2
3
1
2
X
N
3
2
3
M
Ns
X0 ,Y0 ,Z0
Fig.1.1
The separate point in grid can be described by vector of geology-geophisical digital signs, which maximum
dimensionality is limited (maximum 128 signs).
For identification of the profiles and pickets in grid are used their sequence numbers. Herewith the uppermost
profile of the layer always has number 1, either as extreme-left picket on profile.
Any grid in the database can be described by free-form additional text information, containing not more than 56
symbols (for example area or method name). Additional information is asked by all programme modules which
form (as a result of work) new grids can be filled or disregard. So, this is optional information, but its presence
helps at identifications of the grid.
All enumerated parameters of grids are defined by user in time of creation, but there is a possibility to edit existing
grids too.
The file system of “COSCAD 3Dt” database is built to minimize the average access time to data for computing and
service procedures. Herewith quick access is provided to information referring to concrete profile or layer of a grid
owing to increase of access time to information on the same-name pickets of the different profiles or layers in one
grid. That allows optimizing the process of the access to information for the functional procedures of the program
complex. Also structure of the database provides very quick access to separate sign of the grid owing to increase of
access time to vector of signs in separate point of the grid.
It is important that the orientation of processing procedures of the complex on analysis data organized in threedimensional grids does not exclude efficient information handling organized in two-dimensional and onedimensional grids (areal mapping and profiles observations).
The software of the complex includes six functional blocks (modules) uniting procedures on nature of the problems
solved with their help.
For most processing programs of the complex "COSCAD 3Dt" the sign value of grid or several grids are input
information. Such grids and signs are called input data (input grid or input sign). The grids that are formed as a
result of module working are identified as output data (output grid).
After successful installation of the complex "COSCAD 3Dt" it is necessary to create one or several work
directories on hard disk for information accommodation. In each directory it is possible to keep up to 99 grids. It is
not recommended to place data in directory where the software is situated. With information keeping in each
directory it is possible to work using software of the complex keeping in chosen during installation directory (for
instance in C:\COSCAD3Dt).
In the main program window numbers of the networks in databank are flashed (herewith if network is exist a button
corresponding to her is active, otherwise a button is disarmed). Choosing the button of existing grid it is possible to
get grid’s information on parameters, edit the separate parameters and (in case of necessary) delete the grid.
Required functioning program block is chosen from the main menu whereupon appears the program list containing
names of the selected block functional modules. After choice of the module the dialog box with a list of start
parameters will appear in the right part of the screen. Start parameters is identified as information sent to the
processing program for its correct execution (as example, number of input an output grids, number of the processed
sign, value of the key parameter of the program and etc.). For detailed information on start parameters of the
Copyright © 2003 MSGPU
Coscad 3Dt
7
module it is possible (at moment of their filling) to use the help (using "Help" button). Editing of start parameters is
realized by keyboard keys and mouse cursor and ends when the "Start" button is pressed. Whereupon the window
with messages from executing program is appeared. If as a result of working of the program the new grid is created
then before termination of the messages box user will be offered enter additional symbol information about output
grid.
All programs of the complex of spectral-correlation data analysis "COSCAD 3Dt" are divided into six sections:
service, visualization, statistics, filtering, detection and complex. In the name of each section the information about
nature of the problems solved by means of falling modules is kept.
SERVICE
The programs given in this section are intended for database administration and provide some standard data
management system functions. With help of these programs you can realize input/output of information,
association and fragmentation of the grids, filling unmeasured points of the sign, interpolation and extrapolation of
the grid, different data transformations etc.
VISUALIZATION
The complex of spectral-correlation data analysis "COSCAD 3Dt" is equipped by suitable graphic interface
allowing to view 1D, 2D and 3D information from the database on the screen in the manner of separate graphs,
raster maps, cubes of data etc. Except visualization programs the program for decision of the inverse problem of
gravity-magnetic survey enters in the graphic block. Detailed description of these modules you can find in chapter
“Visualization”.
STATISTICS
The programs given in this section are intended for calculation statistical, spectral and correlation descriptions of
geophisical information. The analysis of these features allows to get additional useful information on target field
and it helps to choose the earl of the further processing. The programs falling into group "Sounding" and
"Estimation of parameters of anomaly objects" allow to evaluate the parameters of anomaly by statistical methods.
FILTERING
In programs of this section marketed the most wide-spread in exploratory geophysics linear optimum filters
allowing to solve the tasks of field decomposition, trend removal, estimations of the form of weak anomaly. Special
interest presents the unique adaptive filters allowing correct processing of the nonstationary on spectral-correlation
features geophisical fields.
DETECTION
With the help of these programs the user can solve problems of separation of weak anomaly commensurable on
amplitude with level of the noise (with linear and isometric form) using one or several signs.
COMPLEX
Using the programs of this section allows to solve the problems of the partition of the analyzed area on uniform
areas (classes) with equal average signs value, discernment of complex anomaly on reference anomaly. Besides it is
possible to undertaking component-analysis of a multi-signs data. As input information for programs of this group
the values of different geology-geophisical signs and their derived gotten by means of programs from the other
sections of the complex can be used.
We wish you success in work!
Copyright © 2003 MSGPU
Coscad 3Dt
8
1.2. About Authors
Moscow State Geology Prospecting University
Professor,
Doctor of physico-mathematical sciences,
Member of the Russian Academy of Natural Sciences
A.A. Nikitin
Professor,
Doctor of physico-mathematical sciences,
A.V. Petrov
Senior staff scientist,
A.S. Aleksashin
Coscad 3Dt is Copyright © 1998-2003, MSGPU, All Rights Reserved. Moscow 2003
Copyright © 2003 MSGPU
Coscad 3Dt
9
1.3. How to install "Coscad 3Dt"?
The installation of the complex of spectral-correlation data analysis "COSCAD 3D" includes the following steps:
1. Put the installation CD-disc into the reading device of your computer;
2. Start the setup module COSCAD3D.EXE. (The installation password is “ANDR2001”);
3. After successful closedown of the module COSCAD3D.EXE, the directory C:\COSCAD3D (or other, chosen
by you) will be created. In this directory the special security file “SEC.BIN” will be created.
4. Before start working with program complex you have to send us this file on E-mail: Petrov@msgpa.msgpa.ru or
AlexPetrov76@mail.ru with mark in subject: “SEC.BIN file from …(put you company name here)”;
5. Within 3 days you get the file “NEWSEC.BIN” which necessary to copy in the directory C:\COSCAD3D (or
other, chosen by you at installation). You need this file for correct functioning of our program. For each
computer you need a unique security file;
6. After that you copy of program will be ready to work.
With any questions apply to developers on the following address:
RUSSIA, MOSCOW, Mikluho-Maklaya str., 23, Department of “ЯРМ и ГИ”, room 6-16, professor A.A. Nikitin
OR
by E-Mail: Petrov@msgpa.msgpa.ru or AlexPetrov76@mail.ru
OR
by telephones: (095) 420-51-47, (095) 238-21-95 call for Alex Petrov
fax: (095) 438-14-38
Yours faithfully, Alex Vladimirovich Petrov!
P.S. (ATTENTION!!!): Any attempts of the goal-directed breaking in uncomplicated but multifunctional program
protection can bring to misbehaviour of your operating system and destruction of FAT system at your hard disk.
Besides, computer piracy is pursued under the law.
Copyright © 2003 MSGPU
Coscad 3Dt
10
1.4. FAQ and Error's handling
FREQUENTLY ASKED QUESTIONS LIST:
Q: In case of some visualization modules starting appears the message that DLL is absent.
A: In this case address to the developer.
Q: After starting of the calculation program the message that percent of the execution is a zero on screen is
saved very long time.
A: Restart the calculation program. If the situation is repeated address to the developer.
Q: If the source grid contains the null-codes always appears the message "Unknown error".
A: Most computing modules are oriented on work just with filled grids.
Q: At appearance of the message "Module is absent in given version" at start of all programs address to the
developer.
A: Maybe you did not register your copy of program. Send us the “SEC.BIN” security file.
Please, do not start simultaneously more than five functional accounting modules. This can cause the
deceleration in the program and in your operating system.
ERROR'S HANDLING
(E – Error, D – Error description, S - suggestions)
E: When work with file of the exchange “INREPORT”
D: The mistake of the operating system or low disk space.
S: Try to start the program again or check presence of the free space on your disk.
E: Out of memory
D: The grid is too large for processing by this module or too many applications are working simultaneously.
S: The actions are obvious. Unload some applications and rerun the program.
E: In work with file of the input data
D: The error appears at operations of the institution to information from file.
S: Check correctness of the data in file and their correspondence to description of the format.
E: In work with file of the output data
D: The error appears most often at export of given data into the other formats and is connected with absence of free
disk space.
S: Try to enlarge the available disc space.
E: The window size is less than zero
D: Negative value of one of the window size parameters.
S: Correct the window size.
E: Incorrect number of points of irregular grid
D: The Number of points in file containing information on irregular grid does not comply with number of points
given at the start of the program "Institution of irregular grid".
S: Define the amount of points in irregular grid. Repeat calling of the program module "Institution of irregular
grid".
E: Incorrect geometric parameters of the grid OR Incorrect size of output grid
D: The error appears in the procedure of the institution of the information in the database and is connected with
incorrect geometric parameters of given grid (N, M, Ns, Dx, Dy, Dz etc).
Copyright © 2003 MSGPU
Coscad 3Dt
11
S: Correct the parameters of the source grid.
E: Incorrect factors A, B or inadmissible values of the field
D: In the procedure "Different transformations with data" factors A or B is given incorrect that does not allow to
realize the selected transformation.
S: Correct the factor’s values.
E: Source grids do not coincide on size
D: Some modules require that sizes of two input grids must have the same number of pickets, profiles and layers.
S: Choose the input grids with the same size.
E: The sizes of given matrixes are too small for DACF calculation. (Double-dimensional AutoCorrelation
Function)
D: For DACF calculation it is necessary that sizes of the grid satisfied the following requirements: number of
pickets more than 7, profiles > 3 (if it is chose processing by layers) or layers > 3 (if it is chose processing by
profiles).
S: Choose the source grid with greater sizes or interpolate the grid.
E: The sizes of given matrixes are too small for CCF calculation. (Cross-Correlation Function)
D: For CCF calculation it is necessary that sizes of the grid satisfied the following requirements: number of pickets
more than 5, profiles > 2 (if it is chose processing by layers) or layers > 2 (if it is chose processing by profiles).
S: Choose the source grid with greater sizes or interpolate the grid.
E: The sizes of given matrixes are too small for calculation
D: Some modules require the restriction on minimum number of pickets, profiles and layers of the input grid.
S: Interpolate the given grid with enlarging number of pickets, profiles or layers.
E: The window size is too small. OR The window size is less than 3
D: Window sizes are very small for functioning of the given module.
S: Enlarge window size.
E: Incorrect sizes of the grid
D: Sizes of the grid are very small for functioning of the given module.
S: Interpolate the grid.
E: Discrepancy between parameters of the source and anomalous grids
D: The amount of signs at the anomalous grid less than number of analyzed signs of the source grid or given wrong
borders of the master anomaly.
S: Try to elaborate the reasons of the error. Repeat calling the procedure.
E: Incorrect sign’s numbers
D: The widespread mistake appearing when processing multi-signs grids at choice of the number of processed
signs.
S: Try to elaborate the reasons of the error. Repeat calling the procedure.
E: Incorrect number of the input grid
D: Input grid is absent or does not correspond on size for concrete module.
S: Try to elaborate the reasons of the error. Repeat calling the procedure.
E: Low disk space
D: Current disk is full.
S: Enlarge the free disc space.
E: The input grid does not exist
D: Number of input grid is incorrect.
S: Correct the error. Repeat calling the procedure.
Copyright © 2003 MSGPU
Coscad 3Dt
E: Number of output grid is incorrect
D: The output grid is exist already or occurred the system malfunction.
S: Elaborate the mistake. If the result of the repeated start of the procedure will be negative terminate the work
with complex "COSCAD 3Dt" and repeat its call.
E: The amount of layers or profiles of input grid is less than 3
D: The error meets just when calling the procedures of the three-dimensional grid analysis.
S: Check that grid is three-dimensional and repeat calling the program module.
E: The output grid is exist already
D: The output grid is exist already or occurred the system malfunction.
S: Elaborate the mistake. If the result of the repeated start of the procedure will be negative terminate the work
with complex "COSCAD 3Dt" and repeat its call.
E: Incorrect amount of output grid signs
D: The amount of signs in the output grid less than 1 or more than 128.
S: Correct the error. Repeat calling the procedure.
E: Error in work with temporary file
D: Low free disk space.
S: Enlarge the free disc space.
E: Incorrect Dx, Dy or Dz of the output grid
D: The distance between nearby pickets, profile or layers is less than zero.
S: Correct the error. Repeat calling the procedure.
E: Selected layer or profile is absent in the grid
D: The layer or profile is incorrect for processing.
S: Choose existing in the grid layer or profile.
E: Error! The input grid is a cube or profile
D: This module works just with two-dimension grids.
S: If you want to process the profile convert it into layer with the program "Rotation of the grid".
E: In the input grid Dx is not equal to Dy
D: For given program it is necessary that input grid has to be regular i.e. Dx=Dy.
S: Convert the grid into regular using programs from section "Interpolation".
!!! In case of message "Unknown error" please send the letter to alexpetrov76@mail.ru with headline "ERROR".
Please, send the detail description of the situation led to error. The correction will be send to you in the shortest
period.
Copyright © 2003 MSGPU
12
Coscad 3Dt
13
2. SERVICE
The programs given in section “SERVICE” are intended for database administration and provide some standard data
management system functions. With help of these programs you can realize input/output of information, association
and fragmentation of the grids, filling unmeasured points of the sign, interpolation and extrapolation of the grid,
different data transformations etc.
2.1. Data input...
The programs of this section are intended for inputting of digital signs measured in points of the three-dimensional
(possible in two-dimensional or profile) regular grids from data-files into the database of the complex. Regular grid
is a grid with determined number of pickets, profiles and layers and constant distance between nearest pickets,
profiles and layers. You can not input grids with less than two pickets.
If your grid with observations is not strictly parallelepiped (regular grid) it is necessary to bring it to regular before
inputting to the database of the complex. So, you should fill the empty elements of your grid by any unique values
(so called the dummy-code). This value (dummy-code) is used in the program of the data filling (in section
"SERVICE") for recovering lacking information.
2.1.1. Inputting data from “COSCAD 3Dt” - formats
The module is intended for entering of information from file.
Input parameters of the module:
Output grid
–
the result grid which will be created in the database of the complex;
Input file
–
the full name (with path) of the input data file.
The module is intended for entering of information from files prepared in the following formats:
1. by profiles, by layers, by signs
It is the format where the data in the file are prepared in symbol type with gap or comma separators in the
following structure:
First line:
0
- < format code>
Second line: N, M, Ns, NK
- <Amount of pickets, Amount of profiles, Amount of layer, Amount of signs>
Third line:
XO, YO, ZO
- <the coordinates of the bottom-left corner of a cube >
Fourth line:
Dx, Dy, Dz
- <the steps between pickets, profiles and layers>
Fifth line:
U1, U2
- <the angels defining position of the grid in space >
Then follows directly digital information from the first picket of the first profile of the first layer of the first sign
(profile by profile, layer by layer). After the information on the first sign directly follows (in the same order) the
information on the second sign etc. The amount of information in the separate lines of the source file is free, but the
information for the next layer has to start with a new line.
2. by points, by profiles, by layers
It is the format where the data in the file are prepared in symbol type with gap or comma separators in the
following structure:
First line:
1
- < format code>
Second line: N, M, Ns, NK
- <Amount of pickets, Amount of profiles, Amount of layer, Amount of signs>
Third line:
XO, YO, ZO
- <the coordinates of the bottom-left corner of a cube >
Fourth line:
Dx, Dy, Dz
- <the steps between pickets, profiles and layers>
Fifth line:
U1, U2
- <the angels defining position of the grid in space >
Then follows directly digital information from the first picket of the first profile of the first layer of the first sign
(profile by profile, layer by layer), but in contrast to the first format for each point (if input information is multisign) all signs of the grid are written immediately. After the first sign directly follows the second sign for the each
point etc. The amount of information in the separate lines of the source file is free, but the information for the next
layer has to start with a new line.
Copyright © 2003 MSGPU
Coscad 3Dt
14
3. by pickets, by layers, by signs
It is the format where the data in the file are prepared in symbol type with gap or comma separators in the
following structure:
First line:
2
- < format code>
Second line: N, M, Ns, NK
- <Amount of pickets, Amount of profiles, Amount of layer, Amount of signs>
Third line:
XO, YO, ZO
- <the coordinates of the bottom-left corner of a cube >
Fourth line:
Dx, Dy, Dz
- <the steps between pickets, profiles and layers>
Fifth line:
U1, U2
- <the angels defining position of the grid in space >
Then follows directly digital information but in contrast to the first format the information in the file is situated not
on profiles, but on pickets – it means that at first all first picket’s values in the layer follows then second pickets
etc. After the information on the first sign directly follows (in the same order) the information on the second sign
etc. The amount of information in the separate lines of the source file is free, but the information for the next layer
has to start with a new line.
4. by points, by pickets, by layers
It is the format where the data in the file are prepared in symbol type with gap or comma separators in the
following structure:
First line:
3
- < format code>
Second line: N, M, Ns, NK
- <Amount of pickets, Amount of profiles, Amount of layer, Amount of signs>
Third line:
XO, YO, ZO
- <the coordinates of the bottom-left corner of a cube >
Fourth line:
Dx, Dy, Dz
- <the steps between pickets, profiles and layers>
Fifth line:
U1, U2
- <the angels defining position of the grid in space >
Then follows directly digital information for the first layer (picket by picket, layer by layer) but in contrast to the
third format the information contains all signs for each point immediately. The amount of information in the
separate lines of the source file is free, but the information for the next layer has to start with a new line.
5. by cubes, by signs
It is the format where the data in the file are prepared in symbol type with gap or comma separators in the
following structure:
First line:
4
- < format code>
Second line: N, M, Ns, NK
- <Amount of pickets, Amount of profiles, Amount of layer, Amount of signs>
Third line:
XO, YO, ZO
- <the coordinates of the bottom-left corner of a cube >
Fourth line:
Dx, Dy, Dz
- <the steps between pickets, profiles and layers>
Fifth line:
U1, U2
- <the angels defining position of the grid in space >
This data format is the same with the first format (by profiles, by layers, by signs), but it allows to speed up the
process of the inputting of the information for small grids. If the size of the grid is very large it can occurs errors
connected with shortage of the operative memory which does not appears when you use formats described above.
The amount of information in the separate lines of the source file is free. The requirement that information for the
each layer has to start with a new line does not spread on this format.
6. Archive COSCAD 3Dt format
It is an original software format. The main purpose of this format is the data archiving from the database of the
complex in the suitable structure to quick reconstruction in the base. Archiving is realized by the "Outputting data
in COSCAD 3Dt format".
Copyright © 2003 MSGPU
Coscad 3Dt
15
2.1.2. Inputting data from SURFER format
The module is intended for entering of information from the data file.
Input parameters of the module:
Output grid
–
the result grid which will be created in the database of the complex;
Input file
–
the full name (with path) of the input data file.
The module is intended for input of information from the file prepared in the data format accepted in the program
SURFER 6, SURFER 7 or SURFER 8 for keeping of the coarse observations (so called GRIDs). The file can have
binary or text structure.
2.1.3. Inputting data from GEOSOFT format
The module is intended for entering of information from the data file.
Input parameters of the module:
Output grid
–
the result grid which will be created in the database of the complex;
Input file
–
the full name (with path) of the input data file.
The module is intended for input of information from the file prepared in the data format accepted in the computer
technology GEOSOFT for keeping of the coarse observations (so called GRIDs).
2.1.4. Inputting data from SEGY format
The module is intended for entering of information from the file with data.
Input parameters of the module:
Output grid
–
the result grid which will be created in the database of the complex;
Input file
–
the full name (with path) of the input data file.
The module is intended for input of information from the file prepared in the widespread format of keeping 2D
seismic information - SEGY. The module works just with I1, I2, I4, R4 formats in IBM standard.
2.1.5. Inputting data from INTEGRO format
The module is intended for entering of information from the file with data.
Input parameters of the module:
Output grid
–
the result grid which will be created in the database of the complex;
Input file
–
the full name (with path) of the input data file.
The module is intended for input of information from the file prepared in the data format accepted in the computer
technology INTEGRO for keeping of the coarse observations (so called TOS-format).
2.1.6. Inputting data from RADAR format
The module is intended for entering of information from the file with data.
Input parameters of the module:
Output grid
–
the result grid which will be created in the database of the complex;
Input file
–
the full name (with path) of the input data file.
Operating mode – the format type I2 or I4
The module is intended for input of information from the file prepared in the data format accepted in the computer
technology RADAR.
Copyright © 2003 MSGPU
Coscad 3Dt
16
2.2. Data output...
Modules of the given section are intended for output of information from the database of the complex "COSCAD
3Dt" in the file on your hard disk for further keeping and using in other systems.
2.2.1. Outputting data into COSCAD 3Dt format
The module is intended for output of information from the database into the file.
Input parameters of the module:
Input grid
–
the grid which will be used as a source of information to create the output file;
Input sign
–
the number of the outputting sign.
Format code –
one of the following format codes.
The module is intended for information output from the database into the file with one of the following formats:
1. by profiles, by layers, by signs
It is the format where the data in the file are prepared in symbol type with gap or comma separators in the
following structure:
First line:
0
- < format code>
Second line: N, M, Ns, NK
- <Amount of pickets, Amount of profiles, Amount of layer, Amount of signs>
Third line:
XO, YO, ZO
- <the coordinates of the bottom-left corner of a cube >
Fourth line:
Dx, Dy, Dz
- <the steps between pickets, profiles and layers>
Fifth line:
U1, U2
- <the angels defining position of the grid in space >
Then follows directly digital information from the first picket of the first profile of the first layer of the first sign
(profile by profile, layer by layer). After the information on the first sign directly follows (in the same order) the
information on the second sign etc. The amount of information in the separate lines of the source file is free, but the
information about each layer is started with a new line.
2. by points, by profiles, by layers
It is the format where the data in the file are prepared in symbol type with gap or comma separators in the
following structure:
First line:
1
- < format code>
Second line: N, M, Ns, NK
- <Amount of pickets, Amount of profiles, Amount of layer, Amount of signs>
Third line:
XO, YO, ZO
- <the coordinates of the bottom-left corner of a cube >
Fourth line:
Dx, Dy, Dz
- <the steps between pickets, profiles and layers>
Fifth line:
U1, U2
- <the angels defining position of the grid in space >
Then follows directly digital information from the first picket of the first profile of the first layer of the first sign
(profile by profile, layer by layer), but in contrast to the first format for each point all signs of the grid are written
immediately. After the first sign directly follows the second sign for the each point etc. The amount of information
in the separate lines of the source file is free, but the information about each layer is started with a new line.
3. by pickets, by layers, by signs
It is the format where the data in the file are prepared in symbol type with gap or comma separators in the
following structure:
First line:
2
- < format code>
Second line: N, M, Ns, NK
- <Amount of pickets, Amount of profiles, Amount of layer, Amount of signs>
Third line:
XO, YO, ZO
- <the coordinates of the bottom-left corner of a cube >
Fourth line:
Dx, Dy, Dz
- <the steps between pickets, profiles and layers>
Fifth line:
U1, U2
- <the angels defining position of the grid in space >
Then follows directly digital information but in contrast to the first format the information in the file is situated not
on profiles, but on pickets – it means that at first all first picket’s values in the layer follows then second pickets
etc. After the information on the first sign directly follows (in the same order) the information on the second sign
Copyright © 2003 MSGPU
Coscad 3Dt
17
etc. The amount of information in the separate lines of the source file is free, but the information about each layer is
started with a new line.
4. by points, by pickets, by layers
It is the format where the data in the file are prepared in symbol type with gap or comma separators in the
following structure:
First line:
3
- < format code>
Second line: N, M, Ns, NK
- <Amount of pickets, Amount of profiles, Amount of layer, Amount of signs>
Third line:
XO, YO, ZO
- <the coordinates of the bottom-left corner of a cube >
Fourth line:
Dx, Dy, Dz
- <the steps between pickets, profiles and layers>
Fifth line:
U1, U2
- <the angels defining position of the grid in space >
Then follows directly digital information for the first layer (picket by picket, layer by layer) but in contrast to the
third format the information contains all signs for each point immediately. The amount of information in the
separate lines of the source file is free, but the information about each layer is started with a new line.
5. by cubes, by signs
It is the format where the data in the file are prepared in symbol type with gap or comma separators in the
following structure:
First line:
4
- < format code>
Second line: N, M, Ns, NK
- <Amount of pickets, Amount of profiles, Amount of layer, Amount of signs>
Third line:
XO, YO, ZO
- <the coordinates of the bottom-left corner of a cube >
Fourth line:
Dx, Dy, Dz
- <the steps between pickets, profiles and layers>
Fifth line:
U1, U2
- <the angels defining position of the grid in space >
This data format is the same with the first format (by profiles, by layers, by signs), but it allows to speed up the
process of information output for small grids. If the size of the grid is very large it can occurs errors connected with
shortages of the operative memory which does not appears when you use formats described above. The amount of
information in the separate lines of the source file is free.
6. Archive COSCAD 3Dt format
It is an original software format. The main purpose of this format is the data archiving from the database of the
complex in the suitable structure to quick reconstruction in the base. Archiving is realized by the "Outputting data
in COSCAD 3Dt format".
2.2.2. Outputting data into SURFER format
The module outputs the information from the database into the file.
Input parameters of the module:
Input grid
–
the grid which will be used as a source of information to create the output file;
Input sign
–
the number of the outputting sign.
Layer
–
the number of the outputting layer
Profile
–
the number of the outputting profile
!!!
If both previous parameters are zero the first layer of the grid is outputted.
Dummy-code –
It is a code of unmeasurement data with will be changed into dummy-code of the target
format.
Output file
–
the full name of the file to output. You can browse it using the “Browse…” button
Format
–
one of the following format-codes.
BINARY SURFER format. The binary data format accepted in the program SURFER 7,8 for keeping of the
coarse observations (GRIDs). Such format is intended for keeping only two-dimensional data so in this format you
can output just information on one sign of the grid or by one layer or profile.
ASCII SURFER format. The symbol data format accepted in the program SURFER 7,8 for keeping of the coarse
observations (GRIDs). Such format is intended for keeping only two-dimensional data so in this format you can
output just information on one sign of the grid or by one layer or profile.
Copyright © 2003 MSGPU
Coscad 3Dt
2.2.3. Outputting data into GEOSOFT format
The module outputs the information from the database into the file.
Input parameters of the module:
Input grid
–
the grid which will be used as a source of information to create the output file;
Input sign
–
the number of the outputting sign.
Layer
–
the number of the outputting layer
Profile
–
the number of the outputting profile
!!!
If both previous parameters are zero the first layer of the grid is outputted.
Dummy-code –
It is a code of unmeasurement data with will be changed into dummy-code of the target
format.
Output file
–
the full name of the file to output. You can browse it using the “Browse…” button
Geosoft format –
the format of the GEOSOFT system for keeping 2D information
2.2.4. Outputting data into SEGY format
The module outputs the information from the database into the file.
Input parameters of the module:
Input grid
–
the grid which will be used as a source of information to create the output file;
Input sign
–
the number of the outputting sign.
Layer
–
the number of the outputting layer
Profile
–
the number of the outputting profile
!!!
If both previous parameters are zero the first layer of the grid is outputted.
Dummy-code –
It is a code of unmeasurement data with will be changed into dummy-code of the target
format.
Output file
–
the full name of the file to output. You can browse it using the “Browse…” button
Format code –
the SEGY format with IMB(I2) or IBM(R4) standard.
During the process of the output it is possible to conduct online correction by some flaps of the final SEGY file.
!?! Do not forget to convert the seconds into milliseconds (it is required by the format).
2.2.5. Outputting data into INTEGRO format
The module outputs the information from the database into the file.
Input parameters of the module:
Input grid
–
the grid which will be used as a source of information to create the output file;
Input sign
–
the number of the outputting sign.
Layer
–
the number of the outputting layer
Profile
–
the number of the outputting profile
!!!
If both previous parameters are zero the first layer of the grid is outputted.
Dummy-code –
It is a code of unmeasurement data with will be changed into dummy-code of the target
format.
Output file
–
the full name of the file to output. You can browse it using the “Browse…” button
INTEGRO format – the format of the system INTEGRO for keeping 2D information (so called TOS).
2.2.6. Outputting data into GRAPHER format
The module outputs the information from the database into the file.
Input parameters of the module:
Input grid
–
the grid which will be used as a source of information to create the output file;
Input sign
–
the number of the outputting sign.
Copyright © 2003 MSGPU
18
Coscad 3Dt
19
–
–
!!!
Dummy-code –
Layer
Profile
the number of the outputting layer
the number of the outputting profile
If both previous parameters are zero the first layer of the grid is outputted.
It is a code of unmeasurement data with will be changed into dummy-code of the target
format.
Output file
–
the full name of the file to output. You can browse it using the “Browse…” button
Grapher format – it is intended for output of the one-dimension realizations being the result of the profile and
the layer crossing into the popular system Grapher (the text file X,Y,F).
2.2.7. Outputting data as a common text table X,Y,Z,F (for EXCEL)
The module outputs the information from the database into the file.
Input parameters of the module:
Input grid
–
the grid which will be used as a source of information to create the output file;
Input sign
–
the number of the outputting sign.
Output file
–
the full name of the file to output. You can browse it using the “Browse…” button
EXCEL format –
in the tabular text format are outputted coordinates of all points of the processed sign in the
grid (X,Y,Z,F). The file of such format is easy to load from EXCEL.
2.3. Irregular grids input...
Modules of this section are intended for input of any digital geology-geophisical information measured in points of
the irregular grid of observations into the database of the complex.
2.3.1. Inputting of 2D irregular grids
The module inputs any digital geology-geophisical information measured in points of the two-dimensional irregular
grid of observations. In this module the algorithm of searching for the nearest point in the moving window is used.
Input parameters of the module:
You have give all coordinates and steps in metres.
Output grid
–
the number of the result grid which will be created in the database of the complex;
Minimum coordinate X
–
the minimum coordinate X of the formed regular grid;
Maximum coordinate X
–
the maximum coordinate X of the formed regular grid;
Minimum coordinate Y
–
the minimum coordinate Y of the formed regular grid;
Maximum coordinate Y
–
the maximum coordinate Y of the formed regular grid;
!!!
Attention! You may leave minimum and maximum coordinates with zero
values. In that case all border coordinates will be chosen automatically on the
basis of the analysis of the input information.
Step for X
–
the distance (Dx) between pickets of the created regular grid;
Step for Y
–
the distance (Dy) between profiles of the created regular grid;
!!!
Attention! You may put zero values of the step. In that case this parameter will
be chosen automatically.
Size of the window on X
–
the maximum removing at the searching for the nearest point on X-axis;
Size of the window on Y
–
the maximum removing at the searching for the nearest point on Y-axis;
!!!
Attention! If this parameter is a zero then restriction is not taken, but if it is -1
so the restriction is chosen automatically.
Number of averaging points –
minimum amount of nearest points. Usually this parameter will assign from 3
up to 25. The bigger this parameter the more time of the functioning of the
program and more smoothing the result.
Copyright © 2003 MSGPU
Coscad 3Dt
20
–
the full name (with path) of the input file with irregular grid.
–
having regard to low-frequency components;
–
having regard to mid-range-frequency components;
–
having regard to high-frequency components;
–
taking into consideration just high-frequency components;
The selected method has influence upon the result of the adduction of data to the regular grid. In the first case some
low-frequency components of the given data will be reflected in the resulting grid, in the second case - mid-rangefrequency components will be reflected and etc. It is recommended to conduct the operation of the reduction by
different methods and then choose the best one. In the same way it is possible to use the module "Inputting of 2D
irregular grids (2D spline)" or similar procedures from the other systems (for example, SURFER).
Mode of work
–
allows to smooth the final result. It is recommended.
Input file
Method
The input text file must consist of three values in each line - X, Y and F separated by gaps or commas. X, Y - point
coordinates of two-dimensional irregular grid and F is a value of the sign in this point (first five lines in this file
may have any free text information).
As a result of the program in the database of the complex the regular two-dimensional grid with fixed amount of
pickets and profiles is formed. If the window size is small the created by the program regular grid can contain the
dummy-code. In this case it is necessary to enlarge the window size or using the program "Filling of absent data"
from the section “SERVICE” fill them out. After it (for better smoothing of information) it is possible to use the
module "Adaptive filtering in the window of the alive form" from the section "FILTERING" with small basic
window (3х3) for improvement quality of the filling process.
Working with the start parameter window (located in the right part) it is possible to examine the point’s location of
your irregular grid and choose the borders for future regular grid.
2.3.2. Inputting of 2D irregular grids (2D-spline)
The module inputs any digital geology-geophisical information measured in points of the two-dimensional irregular
grid of observations. This module uses the algorithm of 2D spline-interpolation.
Input parameters of the module:
You have give all coordinates and steps in metres.
Output grid
–
the number of the result grid which will be created in the database of the complex;
Minimum coordinate X
–
the minimum coordinate X of the formed regular grid;
Maximum coordinate X
–
the maximum coordinate X of the formed regular grid;
Minimum coordinate Y
–
the minimum coordinate Y of the formed regular grid;
Maximum coordinate Y
–
the maximum coordinate Y of the formed regular grid;
!!!
Attention! You may leave minimum and maximum coordinates with zero
values. In that case all border coordinates will be chosen automatically on the
basis of the analysis of the input information.
Step for X
–
the distance (Dx) between pickets of the created regular grid;
Step for Y
–
the distance (Dy) between profiles of the created regular grid;
!!!
Attention! You may put zero values of the step. In that case this parameter will
be chosen automatically.
Input file
–
the full name (with path) of the input file with irregular grid.
Mode of work
–
allows to smooth the final result. It is recommended.
The input text file must consist of three values in each line - X, Y and F separated by gaps or commas. X, Y - point
coordinates of two-dimensional irregular grid and F is a value of the sign in this point (first five lines in this file
may have any free text information).
As a result of the program in the database of the complex the regular two-dimensional grid with fixed amount of
pickets and profiles is formed.
Working with the start parameter window (located in the right part) it is possible to examine the point’s location of
your irregular grid and choose the borders for future regular grid.
Copyright © 2003 MSGPU
Coscad 3Dt
21
2.3.3. Inputting of 3D irregular grids
The module inputs any digital geology-geophisical information measured in points of the three-dimensional
irregular grid of observations. In this module the algorithm of searching for the nearest point in the moving window
is used.
Input parameters of the module:
You have give all coordinates and steps in metres.
Output grid
–
the number of the result grid which will be created in the database of the complex;
Size of the window on X
–
the maximum removing at the searching for the nearest point on X-axis;
Size of the window on Y
–
the maximum removing at the searching for the nearest point on Y-axis;
Size of the window on Z
–
the maximum removing at the searching for the nearest point on Z-axis;
!!!
Attention! If this parameter is a zero then restriction is not taken, but if it is -1
so the restriction is chosen automatically.
Minimum coordinate X
–
the minimum coordinate X of the formed regular grid;
Maximum coordinate X
–
the maximum coordinate X of the formed regular grid;
Minimum coordinate Y
–
the minimum coordinate Y of the formed regular grid;
Maximum coordinate Y
–
the maximum coordinate Y of the formed regular grid;
Minimum coordinate Z
–
the minimum coordinate Z of the formed regular grid;
Maximum coordinate Z
–
the maximum coordinate Z of the formed regular grid;
!!!
Attention! You may leave minimum and maximum coordinates with zero
values. In that case all border coordinates will be chosen automatically on the
basis of the analysis of the input information.
Step for X
–
the distance (Dx) between pickets of the created regular grid;
Step for Y
–
the distance (Dy) between profiles of the created regular grid;
Step for Z
–
the distance (Dz) between layers of the created regular grid;
!!!
Attention! You may put zero values of the step. In that case this parameter will
be chosen automatically.
Number of averaging points –
minimum amount of nearest points. Usually this parameter will assign from 5
up to 15. The bigger this parameter - the more time of the functioning of the
program.
Input file
–
the full name (with path) of the input file with irregular grid.
Method
–
having regard to low-frequency components;
–
having regard to mid-range-frequency components;
–
having regard to high-frequency components;
–
taking into consideration just high-frequency components;
The selected method has influence upon the result of the adduction of data to the regular grid. In the first case some
low-frequency components of the given data will be reflected in the resulting grid, in the second case - mid-rangefrequency components will be reflected and etc. It is recommended to conduct the operation of the reduction by
different methods and then choose the best one. In the same way it is possible to use the module "Inputting of 2D
irregular grids (2D spline)" or similar procedures from the other systems (for example, SURFER).
Mode of work
–
is not used in the given module.
The input text file must consist of four values in each line - X, Y, Z and F separated by gaps or commas. X, Y, Z point coordinates of three-dimensional irregular grid and F is a value of the sign in this point (the first line in this
file may have text information. Not more than five words).
As a result of the program in the database of the complex the regular two-dimensional grid with fixed amount of
pickets, profiles and layers is formed. If the window size is small the created by the program regular grid can
contain the dummy-code. In this case it is necessary to enlarge the window size or using the program "Filling of
absent data" from the section “SERVICE” fill them out. After it (for better smoothing of information) it is possible
to use the module "3D Entropy filtration" from the section "FILTERING" with small window (5-7 points.
Copyright © 2003 MSGPU
Coscad 3Dt
22
2.4. Interpolation of grids...
Programs of the given section allow to interpolate of the regular grids (without dummy-code) using methods of
linear, one- and two-dimensional spline and Fourier interpolations.
2.4.1. Liner interpolation
The program allows to interpolate values of one or several signs of the grid by pickets, profiles and layers. In the
program the method of linear interpolation is used. The new distances (steps) must be more than zero.
Input parameters of the module:
Input grid
Sign of input grid
New distance between pickets
New distance between profiles
New distance between layers
Output grid
–
–
–
–
–
–
the source grid;
the processed sign of the grid;
the new distance between pickets (Dx) of the output grid;
the new distance between profiles (Dy) of the output grid;
the new distance between layers (Dz) of the output grid;
the grid containing the result of interpolation.
2.4.2. Profile spline interpolation
The program allows to interpolate values of one or several signs of the grid by pickets or by profiles. In the
program the method of spline interpolation is used. The new distances (steps) must be more than zero.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the grid;
New distance between pickets
–
the new distance between pickets (Dx) of the output grid;
Output grid
–
the grid containing the result of interpolation.
Spline type
–
it is possible the interpolation with use of following splines:
- Compute the cubic spline interpolant with the 'not-a-knot' condition;
- Compute the Hermite cubic spline interpolant;
- Compute the Akima cubic spline interpolant;
- Compute a cubic spline interpolant that is consistent with the concavity of the
data;
- Compute the B-spline interpolant with order 2;
- Compute the B-spline interpolant with order 3;
- Compute the B-spline interpolant with order 4;
- Compute the B-spline interpolant with order 5;
- Compute the B-spline interpolant with order 6;
- Evaluate a piecewise polynomial;
- Evaluate a function defined on a set of points using quadratic interpolation.
2.4.3. 2D Spline interpolation
The program allows to interpolate values of one or several signs of the grid by pickets and profiles. In the program
the method of 2D-spline interpolation is used. The new distances (steps) must be more than zero.
Input parameters of the module:
Input grid
–
Sign of input grid
–
New distance between pickets
–
Copyright © 2003 MSGPU
the source grid;
the processed sign of the grid;
the new distance between pickets (Dx) of the output grid;
Coscad 3Dt
New distance between profiles
Output grid
23
–
–
the new distance between profiles (Dy) of the output grid;
the grid containing the result of interpolation.
2.4.4. Profile Fourier interpolation
The program allows to interpolate values of one or several signs of the grid by pickets or by profiles. In the
program the method of Fourier-interpolation is used. The module is very efficient at building of geological cuts.
The new distances (steps) must be more than zero.
Input parameters of the module:
Input grid
Sign of input grid
New distance between pickets
Output grid
–
–
–
–
the source grid;
the processed sign of the grid;
the new distance between pickets (Dx) of the output grid;
the grid containing the result of interpolation.
2.5. Extrapolation of grids
This program allows to extrapolate grids on the along-profile direction and transversely of the profile strike.
Input parameters of the module:
Input grid
Sign of input grid
Amount of points on the left
Amount of points on the right
Amount of points from above
Amount of points from below
Output grid
–
–
–
–
–
–
–
the source grid;
the processed sign of the grid;
the number of extrapolated pickets on the left;
the number of extrapolated pickets on the left;
the number of extrapolated profiles from above;
the number of extrapolated profiles from below;
the grid containing the result of extrapolation.
In the program an original algorithm to extrapolation is used. The good results are got if number of extrapolated
points on the left, on the right, from below and overhand does not exceed 5-15% from the total number of pickets
or profiles of the grid. For exception of edge effect appearing after usage of different modules the preliminary grid
extrapolation is recommended. After undertaking of some calculations the resulting grid can be bring to the source
size with the help of "Grid fragmentation" module.
2.6. Filling of absent data
The program is intended for filling being absent values (so called dummy-code) in the grid. The program uses the
original filling algorithm that does not cause profound changes of the spectral-correlation attributes of the field.
The program works correct if the share of dummy values does not exceed 25%.
Input parameters of the module:
Input grid
Sign of input grid
Filling value
Output grid
Copyright © 2003 MSGPU
–
–
–
–
the source grid;
the processed sign of the grid;
the number complying with the dummy-code of the grid;
the full grid without dummy values.
Coscad 3Dt
24
2.7. Fragmentation of grids
The program makes a new fragment-grid from any existing in the database of the complex grid.
Input parameters of the module:
Input grid
Left picket
Right picket
Upper profile
Bottom profile
Upper layer
Bottom layer
Output grid
–
–
–
–
–
–
–
–
the source grid;
left border for a new grid by pickets;
right border for a new grid by pickets;
upper border for a new grid by profiles;
bottom border for a new grid by profiles;
upper border for a new grid by layers;
bottom border for a new grid by layers;
the grid containing the result of fragmentation.
Amount of signs in the input and output grids are equal. Numbers of the first and last picket and the first and last
profile of the input grid define fragment boundaries.
2.8. Inserting of grids
This program is intended for "insertions" in the grid of the other grid with smaller size. Changing the key parameter
of the method it is possible to simply impose the information being kept in the grid of the smaller size or take the
average between values of the signs in the corresponding points in both grids.
Input parameters of the module:
Input grid
–
the source grid in which the fragment will be inserted;
Sign of input grid
–
the processed sign of the source grid;
Second input grid
–
the grid which will be implanted;
Second sign of input grid
–
the processed sign of the second grid;
Bottom-left picket
–
the picket defining position of the insertion in the source grid;
Upper profile
–
the profile defining position of the insertion in the source grid;
Upper layer
–
the layer defining position of the insertion in the source grid;
Output grid
–
the full grid without dummy values;
Insertion method
–
there are four ways of the fragment insertion:
The First one is to completely substitutes values in the source grid with values of the fragment.
The Second one is to change values of the source grid on to average between signs of both source and
fragment grids.
The Third one allows to realize the operation of the insertion with provision for dummy-code. It means
that if the dummy-code (131313Е13) is present in the source grid then this point does
not subject to operation of the insertion.
The Last one allows filling the certain fragment by the dummy-code (131313E13).
The operation of the insertion reasonable to conduct just after detailed analysis of the fragment or if the certain area
is required to be filled by the dummy-code.
2.9. Rotating of grids
The program allows conducting the different geometric transformations with location of pickets, profiles and layers
of the source grid.
Input parameters of the module:
Copyright © 2003 MSGPU
Coscad 3Dt
25
Input grid
–
the source grid;
Output grid
–
the rotated grid;
Rotation direction
–
the way of the rotation.
The last parameter defines concrete transformation method:
1. The rotation of the grid in plane of the profiles and pickets from right to left. The program realizes exactly
tumbling of the source grid but not the transposition.
2. The rotation of the grid in plane of the profiles and pickets from left to right. The program realizes exactly
tumbling of the source grid but not the transposition.
3. Changing lower profiles onto upper profiles.
4. Changing left pickets onto right pickets.
5. Changing lower layers onto upper layers.
6. Changing profiles onto layers.
7. Changing layers onto profiles.
2.10. Uniting of grids
This program allows uniting some signs of different grids into one general grid. It is expected that sizes of the input
grids coincide otherwise the size of the resulting grid will be determined by minimum number of pickets, profiles
and layers out of two grids.
Input parameters of the module:
Input grid
–
the first uniting grid;
Second input grid
–
the second uniting grid;
Third input grid
–
the third uniting grid;
Fourth input grid
–
the fourth uniting grid;
Fifth input grid
–
the fifth uniting grid;
Output grid
–
the result grid;
The amount of the united grids can not be more than five. The concrete amount of the united grids is defined by the
following rule: If, for example, the number of the third source grid equals zero then just first and second source
grids will be united. The amount of united signs in each grid and their numbers are defined in the dialogue after
starting of the program. The amount of signs in the resulting grid equals to total of chosen signs in all input grids.
Most often the program is used before using the algorithm of multi-signs analysis of information from the section
"COMPLEX" for which is necessary that all analyzed signs were found in one source grid.
2.11. Some transformations of data
Different algebraic operations with one or two signs of the grid are provided in the program complex. Except
algebraic transformations with data the module allows to change one definite value of the sign onto any another
assigned by user. For example, it is possible to create the grid containing dummy-code in determined points from
existing in the database grid (so called mask-grid) with dummy-code. Creation of the grid with dummy-code is
necessary when you want to remove the results of the calculations from the grid originally putted in the database
with dummy-code.
2.11.1. Transformations with one sign
This module is intended for realization of the algebraic transformations with one sign of the grid.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Output grid
–
Copyright © 2003 MSGPU
the source grid;
the processed sign of the source grid;
the result grid;
Coscad 3Dt
Coefficient A
Coefficient B
Coefficient C
Transformation type
26
–
the value of factor A of chosen transformation;
–
the value of factor B of chosen transformation;
–
the value of factor C of chosen transformation;
–
one of the following types:
1. Liner
Ax+B;
2. Inverse
A/(x-B);
3. Logarithmical Aln(x);
4. Exponential
Aexp(x);
5. Power
Ax^B;
6. Changing the values equals to A onto C;
7. Changing the values greater than A onto C;
8. Changing the values smaller than A onto C;
9. Absolute value |x|;
10. Projection of grid values onto range [A,B]. It means that it is used the
common linear transformation allowing to bring size of parameter changing
of given data into given interval.
2.11.2. Transformations with two signs
This module is intended for realization of the algebraic transformations with two signs of one or two grids.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the first source grid;
Second input grid
–
the second grid (optional);
Sign of second input grid
–
the processed sign of the second source grid (optional);
Coefficient A
–
the value of factor A of chosen transformation;
Coefficient B
–
the value of factor B of chosen transformation;
Output grid
–
the result grid;
Transformation type
–
one of the following types:
1. Sum
Ax + By;
2. Product
Ax * By;
3. Division
Ax / By;
4. If in the second grid “A” is kept then on this place of the first grid will be “B”.
2.11.3. Imposition of dummy-code
This module is intended for imposition of the dummy-code in the grid.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Second input grid
–
the second source grid with dummy-code (so called “mask-grid”);
Sign of second input grid
–
the processed sign of the second source grid;
Dummy-code of input grid
–
the dummy-code of the source grid;
Dummy-code of mask-grid
–
the dummy-code of the second source grid – “mask-grid”;
Output grid
–
the result grid;
The result grid containing dummy-code in some determined points is being built on existing in the database grid
with dummy-code (so called "mask-grid").
Obviously that the grid with dummy-code and the source grid which is superimposed by dummy-code have to be
equal by size (i.e. contain the alike numbers of pickets and profiles). If you had interpolate the grid during
calculation then for reception of a new "mask-grid" (with new interpolated size) you have to interpolate the source
grid containing dummy-code. Creation of the grid with dummy-code is necessary when you want to remove the
results of the calculations from the grid originally putted in the database with dummy-code. For example, to print
the result grid from SURFER.
Copyright © 2003 MSGPU
Coscad 3Dt
27
If the source grid has three-dimensions then as a result of the module functioning you will get the threedimensional grid which will contain the dummy-code in accordance with the "mask-grid" at all layers, herewith the
"mask-grid" has to be two-dimensional.
2.12. Gluing of grids
This program is intended for joining (gluing or splicing) of one sign of the grid with another. Grids can stick
together on profiles, pickets or layers (i.e. with front- or right- or bottom-side of the grid).
Input parameters of the module:
Input grid
Sign of input grid
Second input grid
Sign of second input grid
Output grid
Gluing method
–
–
–
–
–
–
the source grid which will be added by other;
the processed sign of the source grid;
the second source grid that will be added to the source grid;
the processed sign of the second source grid;
the result grid;
the second grid can be glued to the first by deferent ways - to the right,
below and behind.
When gluing it is necessary that component parts having corresponding geometric sizes. So if you are gluing grids
"on the right" it is necessary that number of profiles and layers in both grids coincided. In case of “from below”gluing grids have to be equal by pickets and profiles.
The "Gluing of unequal grids"- mode allows to stick together two-dimensional grids with different situation and
size. In this case the Dx and Dy parameters of sources grids must to be equal.
Grids must to be “comparable”. This means that grid with X-coordinates 100,200,300… can not be glued with grid
having X-coordinates 450,550,650… Therefor for gluing these grids it is necessary that second grid had to have Xcoordinates 400,500,600… or the first grid must have X-coordinates like 150,250,350…
Grids can be both overlap and non-overlap.
The program is used when some additional information on the under investigation territory is appeared.
2.13. Printing information into file
This program is outputting values of one sign of the given grid into the print-file “printer.mga”. This file is
situated in the current directory.
Input parameters of the module:
Input grid
–
the source grid;
Left picket
–
lest border for the fragment to print by pickets;
Right picket
–
right border for the fragment to print by pickets;
Upper profile
–
upper border for the fragment to print by profiles;
Bottom profile
–
bottom border for the fragment to print by profiles;
Upper layer
–
upper border for the fragment to print by layers;
Bottom layer
–
bottom border for the fragment to print by layers;
Language
–
the label’s language (Russian or English).
Information of the print-file can be viewed by "Viewing the print-file "-module from the section "SERVICE".
2.14. Adaptive equalization of seismic traces
This program corrects defective seismic information when on different seismic traces there is different dispersion’s
level.
Input parameters of the module:
Input grid
–
Copyright © 2003 MSGPU
the source grid;
Coscad 3Dt
28
Sign of input grid
–
the processed sign of the source grid;
Output grid
–
the result grid;
Width of base-window for
estimation of dispersion
–
the amount of traces for average level of dispersion estimation. It is being
chosen in accordance with concrete data.
Copyright © 2003 MSGPU
Coscad 3Dt
29
3. VISUALIZATION
The complex of spectral-correlation data analysis "COSCAD 3Dt" is equipped by suitable graphic interface
allowing to view 1D, 2D and 3D information from the database on the screen in the manner of separate graphs,
raster maps, cubes of data etc.
3.1. Raster map
The program allows to view information from the database as common raster map. This module works in multipledocument interface (MDI) allowing view of information in several windows simultaneously.
The grid for viewing is being selected from grid’s catalogue (main program form) before starting the program
"Raster map" or immediately after module start.
The module "Raster map" allows following things:
1. You can open a new window for viewing of the grid. Choose the submenu "Select the grid" from menu
"Service".
2. Change the viewing range. Choose submenu "Viewing range" from "Service" menu. Herewith if the “Fill”mode is selected then data with values smaller than chosen range will be filled with color corresponding to the
minimum value while data with values greater than maximum range-value will be filled with color
corresponding to the maximum value. If “Fill”-mode is not chosen that data out of the chosen range does not
plot at all.
3. Creation, selection and loading of a color-palette. Choose submenu "Palette" from menu "Service". The
dialog of the color-palette choice will be loaded. In this dialog you may choose among existed standard palettes
or build and save your own one. For making of new palette assign several white rectangles located under the
current palette with any colors from the list below. For this fix the rectangle by the mouse cursor and choose
the necessary color with double-click. Assigned several rectangles with colors and choose the button "Build"
for viewing the palette.
4. Changing of axial labels. Choose submenu "Coordinates" from menu "Service". Axes possible to have labels
in pickets and profiles or in real coordinates.
5. Close the current window. Choose item "Close the window" from menu "Service".
6. Choose the concrete sign of the grid for viewing. The point "Sign" of the main menu is using for sign choice.
7. View any cut or layer of any three-dimensional grid. For layer or cut choosing use the point "Layer/Profile"
of the main menu. If the grid is two-dimensional this menu-item is not available.
8. Map scaling. In the current window choose the area of the viewing having pressed left mouse key. After
appearance of the selected area boarded with stroke line press the right button (!!! holding the left button). The
selected area will be boarded with solid line. If you want to correct the fragment repeat actions described above.
Then choose the point "Scale" of the main menu and choose “zoom in”.
9. Build the graphs along the current coordinate line or along any directions. In the current window choose
the spread direction of graphics having pressed left button. After appearance of the selected area boarded with
stroke line holding the left button press the right button (for area fixing). The selected direction (diagonal of the
chosen area) will be plotted with solid line. If you want to correct the direction repeat actions described above.
Then choose the point "Graphs" from the main menu and choose item “Along the direction”. The separate
window for the graphic will be appeared. Using this window you may save the text file with coordinates X,Y
and values of the field F.
10. Plotting of the chosen layer or cut of the grid. Choose the point "Visualization" of the main menu. In some
cases (for example, at the beginning or at opening of a new window) the program expects the choice of the
point "Visualization" before plotting of the raster map.
11. Choosing the current window of disposing of the opened windows on the screen. The submenu "Window"
from the main menu.
12. Get the help. The point "Help" of the main window.
13. To move between layers and profiles use following keys: PageUp, PageDown, Home and End.
14. Edit the separate points in the interactive-mode. Choose the point to edit by mouse or by arrow-keys (Left,
Right,Up,Down). Then press the “Enter” key and enter a new point value. For corrections viewing choose the
Copyright © 2003 MSGPU
Coscad 3Dt
30
point "Visualization" from the main menu. If you want to save changes to the database it is necessary to choose
the point " Save changes" from the section "Service".
15. Filling with the dummy-code of chosen squared area. In the current window choose the squared area
(rectangular fragment) of the graphic having pressed left button. After appearance of the selected area boarded
with stroke line holding the left button press the right button (for area fixing). The selected area will be plotted
with solid line. If you want to correct the direction repeat actions described above. Then press the “Enter” key.
For corrections viewing choose the point "Visualization" from the main menu. If you want to save changes to
the database it is necessary to choose the point " Save changes" from the section "Service".
16. Saving of the current profile or layer into the separate grid of the database or SURFER-format. Use the
corresponding items from the section "Service" from the main menu. If the grid is two-dimensional this item
does not work.
17. Show the field gradient. Choose the point "Gradients" from the section "Service". On the plotted curve (raster
map) lines are appeared. Their directions will coincide with the direction of the full gradient of given field and
thier length will be corresponded with the value of the full gradient.
18. Imposing of points onto the raster map. First of all it is necessary to prepare the text file with lines
containing number triplet separated with gaps – X,Y coordinates and F value in the point. Then choose the
corresponding menu item.
19. Fragment saving allows to save the examined fragment in any chosen grid of the database. It works only when
you see the fragment.
20. Saving with new range. If you had changed the color-range it is possible (without exiting from theraster map
window) to create the grid in which the information within the chosen range will be kept.
Also you can make the screen shot using “Print Screen” key.
During viewing process you may get values of the coordinates of any point and values of the sign in this point. You
can find this information at the bottom information line under the raster map. You just have to choose the point by
the mouse cursor and in the lower part of the screen coordinates X,Y,Z and importance of the sign F will be
flashed.
For calculation of the statistical parameters of the examined fragment choose the point "Statistics” from the section
"Service".
3.2. Plotting of classification
This module allows plotting and editing the grids that are created by classification programs. Besides using this
module it is possible to visualize grids containing directions of anomaly strikes which are got as a result of
following modules usage: the module "Two-dimensional adaptive filtering" from the section "Filtering" and "Selftuning filtering" from the section "Detection".
This module works in multiple-document interface (MDI) allowing view of information in several windows
simultaneously. The grid for viewing is being selected from grid’s catalogue (main program form) before starting
the program "Plotting of classification" or immediately after module start.
The module "Plotting of classification" allows following things:
1. You can open a new window for viewing of the grid. Choose the submenu "Select the grid" from menu
"Service".
2. Changing of axial labels. Choose submenu "Coordinates" from menu "Service". Axes possible to have labels
in pickets and profiles or in real coordinates.
3. Close the current window. Choose item "Close the window" from menu "Service".
4. Change the color palette. Double click on the chosen color from rectangles at the bottom of the screen.
5. Choose the concrete sign of the grid for viewing. The point "Sign" of the main menu is using for sign choice.
6. View any cut or layer of any three-dimensional grid. For layer or cut choosing use the point "Layer/Profile"
of the main menu. If the grid is two-dimensional this menu-item is not available.
7. Map scaling. In the current window choose the area of the viewing having pressed left mouse key. After
appearance of the selected area boarded with stroke line holding the left button press the right button. The
selected area will be boarded with solid line. If you want to correct the fragment repeat actions described above.
Then choose the point "Scale" of the main menu and choose “zoom in”.
8. Build the graphs along the current coordinate line or along any directions. In the current window choose
the spread direction of graphics having pressed left button. After appearance of the selected area boarded with
Copyright © 2003 MSGPU
Coscad 3Dt
31
stroke line holding the left button press the right button (for area fixing). The selected direction (diagonal of the
chosen area) will be plotted with solid line. If you want to correct the direction repeat actions described above.
Then choose the point "Graphs" from the main menu and choose item “Along the direction”. The separate
window for the graphic will be appeared. Using this window you may save the text file with coordinates X,Y
and values of the field F.
9. Plotting of the chosen layer or cut of the grid. Choose the point "Visualization" of the main menu. In some
cases (for example, at the beginning or at opening of a new window) the program expects the choice of the
point "Visualization" before plotting of the raster map.
10. Choosing the current window of disposing of the opened windows on the screen. The submenu "Window"
from the main menu.
11. To move between layers and profiles use following keys: PageUp, PageDown, Home and End.
12. Edit the separate points in the interactive-mode. Choose the point to edit by mouse or by arrow-keys (Left,
Right,Up,Down). Then press the “Enter” key and enter a new point value. For corrections viewing choose the
point "Visualization" from the main menu. If you want to save changes to the database it is necessary to choose
the point " Save changes" from the section "Service".
13. Get the help. The point "Help" of the main window.
14. Filling with the dummy-code of chosen squared area. In the current window choose the squared area
(rectangular fragment) of the graphic having pressed left button. After appearance of the selected area boarded
with stroke line holding the left button press the right button (for area fixing). The selected area will be plotted
with solid line. If you want to correct the direction repeat actions described above. Then press the “Enter” key.
For corrections viewing choose the point "Visualization" from the main menu. If you want to save changes to
the database it is necessary to choose the point " Save changes" from the section "Service".
15. Save current or choose preserved palette. Use the corresponding items from the section "Service" of the
main menu.
16. Build histograms on classes and signs. Choose the point "Building of histograms" from the section "Service".
Then in right part of screen will appear K windows (K – an amount of signs on which the classification was
realized). In each window the sign’s sharing in classes is expressed by colors corresponding to classes on the
screen. It is possible to examine sign’s distribution in all class immediately or apart in each class - refer to point
"Service".
You can use the key “Print Screen” or any screen invader to make a screen shot.
3.3. Graphics
The module "Graphics" builds one or several graphs on the same-name profile from several grids. It is possible to
plot up to 12 different graphs simultaneously. This module works in multiple-document interface (MDI) allowing
view of information in several windows simultaneously.
Module’s menu has following items:
Grid selection
- allows to open a new window for viewing the certain sign or signs of the chosen grid.
Type
- allows to change the thickness of graphic’s line and to switch on/off “draw points” mode.
Close the window
- close the current window.
Visualization
- updates drawings.
Scale
- allows zooming in the picture and drawing graphs in the individual scale. For detailing of
the viewing stretch (holding the left mouse button pressed) rectangle in the field of the plot
and not releasing left button press the right mouse button. For recovering of the initial scale
choose "Source scale" mode.
Add
- allows adding a graph of the certain sign of the chosen grid.
Delete
- if you are examining more than one graphics this function allows to exclude the chosen in
the table graph from viewing area. Select in the table any graph and delete it.
The color rounded button allows to realize dynamic (animated) viewing i.e. sequential viewing of graphs from
profile to profile.
The upwards/downwards arrows allow to change the current profile.
The right/left arrows allow changing graphics to the right or to the left. Motion is possible only if you see just a
fragment of the graph (not all profile).
Getting information about the grid. Press the left mouse button in the corresponding line of the table.
Copyright © 2003 MSGPU
Coscad 3Dt
32
3.4. Inverse problem
The module "Inverse problem" solves the inverse problem of a gravimetric and magnetometric prospecting for
chosen profile of the grid by means of the analytical continuation of the field in the lower half-space up to the depth
of 1/3 lengths of the profile.
Program’s menu has following points:
Grid selection
- allows to open a new window for viewing the certain sign of the chosen grid.
Parameters
- allows to choose the type of the field - magnetic field Z, magnetic field T, gravitational
field and to choose parameters for magnetic field: the angle of full magnetization vector T
with horizon and the angle of the vector of the magnetizing J with horizon.
Visualization
- updates drawings.
The color rounded button allows to realize dynamic (animated) viewing i.e. sequential viewing of graphs from
profile to profile.
The upwards/downwards arrows allow to change the current profile.
The right/left arrows allow changing graphics to the right or to the left. Motion is possible only if you see just a
fragment of the graph (not all profile).
Pressing left mouse button on the cut you may get the value of the depth of the current point in units of the
measurement between pickets.
3.5. Viewing in projections X, Y, Z
The module "3D raster cube" allows to display 3D data of the format of "COSCAD 3Dt" technology in the manner
of a raster cube in projections X,Y,Z and to slice any parallelogram-fragment of the cube.
The main window contains following menu items:
Menu Service:
Grid selection
- for opening of a new grid;
Points
- plotting of the cube points (assigned in the file coordinates).
Coordinates
- axes possible to have labels in pickets and profiles or in real coordinates.
Menu View:
Control panel
Hide the cursor
Cube boarders
Cursor coordinates
Menu management:
Build the grid
Rebuild/Update
Update all
Cut out
Viewing range
Filter/range
Cuts
- the panel on/off switcher.
- hides the cursor in the visualization window;
- boarder on/off switcher (viewing the cut);
- displays the current coordinates of the mouse cursor in the visualization window;
- builds the cube in raster mode (by default) in accordance with chosen matrix;
- reconstructs the cube reflecting last change made in the active window.
- reconstructs all cubes in all opened window;
- cuts the part of the cube in accordance with selected values of pickets, profiles, layers;
- allows:
- to see the histogram of values by colors;
- to change the color-palette range of the current matrix having chosen a new minimum and
maximum values;
- data out of the range are not drawn / are drawn;
- switch to “Cuts viewing” mode;
The “Cuts viewing” mode:
Palette
- choose or create the palette.
Copyright © 2003 MSGPU
Coscad 3Dt
33
Plane
- choose the direction to viewing planes (projection plane).
Step
- choose the step between plotted profiles.
Growth
- choose the direction of the plane’s moving (on increasing of profiles or otherwise).
Animation
- change the view mode (with conservation of all planes or otherwise).
Play
- start animation.
Pause
- pause animation.
Stop
- finish animation.
When all child windows are locked the menu contains just two items: "Grid selection" and "New window".
Each child window contains:
- a possibility of the matrix choice by the item "Matrix" and "Sign" (the current matrix is
being showed in the window’s headline);
- a possibility to choosing the place of the cut in the cube ("Picket", "Profile", "Layer");
- a possibility of ready palettes choice (only in “raster”-mode);
- a possibility of the building quality choice (as for “raster” so and for “class” mode): high
(Hi), average (Med), low (Low).
In the “raster” mode the legend shows maximum and minimum values of the displayed planes and corresponding
color of the selected palette.
In the “class” mode the legend shows colors of corresponding classes allowing changing the color of any classes
(double click) and to look the amount of values in each class.
The coordinates of the point (the cursor) inside the cube - values were provided corresponding to units of the
measurement and with a provision for coordinates of the left lower corner of the grid.
In the visualization window some actions can be realized by means of keyboard:
Right, Left
- moving between profiles;
Up, Down
- moving between layers;
Shift+Left, Shift+Right - moving along a profile (between pickets);
Enter
- build the cut;
Space
- hide the cursor.
The same actions can be realized with mouse:
Moving between profiles, pickets, layers - drag over the active plain cursor (thick white line) into necessary
point;
Build the cut - a double click on the cube;
Hide the cursor - a click outside of the cube.
3.6. Classification viewing in projections X, Y, Z
The module "Classification viewing in projections X, Y, Z" allows to display 3D data of the format of "COSCAD
3Dt" technology in the manner of a raster cube in projections X,Y,Z and to slice any parallelogram-fragment of the
cube.
The main window contains following menu items:
Menu Service:
Grid selection
- for opening of a new grid;
Points
- plotting of the cube points (assigned in the file coordinates).
Coordinates
- axes possible to have labels in pickets and profiles or in real coordinates.
The screen invader - for conservation of the scene in the buffer (so called “screen shot”).
Menu View:
Control panel
Hide the cursor
Cube boarders
Cursor coordinates
- the panel on/off switcher.
- hides the cursor in the visualization window;
- boarder on/off switcher (viewing the cut);
- displays the current coordinates of the mouse cursor in the visualization window;
Menu management:
Build the grid
- builds the cube in raster mode (by default) in accordance with chosen matrix;
Rebuild/Update
- reconstructs the cube reflecting last change made in the active window.
Copyright © 2003 MSGPU
Coscad 3Dt
Update all
Cuts
Cut out
34
- reconstructs all cubes in all opened window;
- switch to “Cuts viewing” mode;
- cuts the part of the cube in accordance with selected values of pickets, profiles, layers;
The “Cuts viewing” mode:
Palette
- choose or create the palette.
Plane
- choose the direction to viewing planes (projection plane).
Step
- choose the step between plotted profiles.
Growth
- choose the direction of the plane’s moving (on increasing of profiles or otherwise).
Animation
- change the view mode (with conservation of all planes or otherwise).
Play
- start animation.
Pause
- pause animation.
Stop
- finish animation.
When all child windows are locked the menu contains just two items: "Grid selection" and "New window".
Each child window contains:
- a possibility of the matrix choice by the item "Matrix" and "Sign" (the current matrix is
being showed in the window’s headline);
- a possibility to choosing the place of the cut in the cube ("Picket", "Profile", "Layer");
- a possibility of ready palettes choice (only in “raster”-mode);
- a possibility of the building quality choice (as for “raster” so and for “class” mode): high
(Hi), average (Med), low (Low).
In the “raster” mode the legend shows maximum and minimum values of the displayed planes and corresponding
color of the selected palette.
In the “class” mode the legend shows colors of corresponding classes allowing changing the color of any classes
(double click) and to look the amount of values in each class.
In the visualization window some actions can be realized by means of keyboard:
Right, Left
- moving between profiles;
Up, Down
- moving between layers;
Shift+Left, Shift+Right - moving along a profile (between pickets);
Enter
- build the cut;
Space
- hide the cursor.
The same actions can be realized with mouse:
Moving between profiles, pickets, layers - drag over the active plain cursor (thick white line) into necessary
point;
Build the cut - a double click on the cube;
Hide the cursor - a click outside of the cube.
3.7. 3D Surface view
The program allows viewing information from the database as three-dimensional surfaces. This module works in
multiple-document interface (MDI) allowing view of information in several windows simultaneously. The grid for
viewing is being selected from grid’s catalogue (main program form) before starting this program or immediately
after loading.
The module "Raster map" allows following things:
1. You can open a new window for viewing of the grid. Choose the submenu "Select the grid" from menu
"Service".
2. Creation, selection and loading of a color-palette. Choose submenu "Palette" from menu "Service". The
dialog of the color-palette choice will be loaded. In this dialog you may choose among existed standard palettes
or build and save your own one.
3. Close the current window. Choose item "Close the window" from menu "Service".
4. Choose the concrete sign of the grid for viewing. The point "Sign" of the main menu is using for sign choice.
5. View any cut or layer of any three-dimensional grid. For layer or cut choosing use the point "Layer/Profile"
of the main menu. If the grid is two-dimensional this menu-item is not available.
Copyright © 2003 MSGPU
Coscad 3Dt
35
6. Plotting of the chosen layer or cut of the grid. Choose the point "Visualization" of the main menu. In some
cases (for example, at the beginning or at opening of a new window) the program expects the choice of the
point "Visualization" before plotting of the raster map.
7. Choosing the current window of disposing of the opened windows on the screen. The submenu "Window"
from the main menu.
8. Get the help. The point "Help" of the main window.
9. Hide the color scale.
10. Hide axes labels.
11. Change the viewing form.
For using of other possibilities press the right mouse button at the surface plot.
3.8. Graphic’s map
This module allows viewing information from the database in the manner of a map of graphics. This module works
in multiple-document interface (MDI) allowing view of information in several windows simultaneously. The grid
for viewing is being selected from grid’s catalogue (main program form) before starting the program "Graphic‘s
map" or immediately after module start.
The module "Plotting of classification" allows following things:
1. You can open a new window for viewing of the grid. Choose the submenu "Select the grid" from menu
"Service".
2. Change the viewing range. Choose submenu "Viewing range" from "Service" menu.
3. Creation, selection and loading of a color-palette. Choose submenu "Palette" from menu "Service". The
dialog of the color-palette choice will be loaded. In this dialog you may choose among existed standard palettes
or build and save your own one. For making of new palette assign several white rectangles located under the
current palette with any colors from the list below. For this fix the rectangle by the mouse cursor and choose
the necessary color with double-click. Assigned several rectangles with colors and choose the button "Build"
for viewing the palette.
4. Changing of axial labels. Choose submenu "Coordinates" from menu "Service". Axes possible to have labels
in pickets and profiles or in real coordinates.
5. Close the current window. Choose item "Close the window" from menu "Service".
6. Choose the concrete sign of the grid for viewing. The point "Sign" of the main menu is using for sign choice.
7. Map scaling. In the current window choose the area of the viewing having pressed left mouse key. After
appearance of the selected area boarded with stroke line holding the left button press the right button. The
selected area will be boarded with solid line. If you want to correct the fragment repeat actions described above.
Then choose the point "Scale" of the main menu and choose “zoom in”.
8. Build the graphs along the current coordinate line or along any directions. In the current window choose
the spread direction of graphics having pressed left button. After appearance of the selected area boarded with
stroke line holding the left button press the right button (for area fixing). The selected direction (diagonal of the
chosen area) will be plotted with solid line. If you want to correct the direction repeat actions described above.
Then choose the point "Graphs" from the main menu and choose item “Along the direction”. The separate
window for the graphic will be appeared. Using this window you may save the text file with coordinates X,Y
and values of the field F.
9. Plotting of the chosen layer or cut of the grid. Choose the point "Visualization" of the main menu. In some
cases (for example, at the beginning or at opening of a new window) the program expects the choice of the
point "Visualization" before plotting of the raster map.
10. Choosing the current window of disposing of the opened windows on the screen. The submenu "Window"
from the main menu.
11. Get the help. The point "Help" of the main window.
Also you can make the screen shot using “Print Screen” key.
During viewing process you may get values of the coordinates of any point and values of the sign in this point. You
can find this information at the bottom information line under the raster map. You just have to choose the point by
the mouse cursor and in the lower part of the screen coordinates X,Y,Z and importance of the sign F will be
flashed.
Copyright © 2003 MSGPU
Coscad 3Dt
36
4. STATISTICS
The programs given in this section are intended for calculation statistical, spectral, correlation and gradient
descriptions of geophisical information. The analysis of these features allows to get additional useful information
on target field and it helps to choose the earl of the further processing.
Modules of this block are realized first stage of data processing on which following problems are being solved:
 Analysis of statistical and spectral-correlation characteristics of a geophisical field for estimation of parameters
of anomalous forming of a field and complicating hindrances (frequency spectrum, statistical and correlation
characteristics, amplitude, size, directions of anomaly strikes);
 Calculation of statistical characteristics of geophisical fields in so-called "slithering window" and gradient
characteristics which are directly analyzed in the process of the interpretation as well as can be enclosed in
processing of multi-signs data by the algorithms of the artificial perception and classification;
 With provision of estimated statistical and spectral-correlation characteristics the earl of the further processing
of the under investigation geophisical field will have been chosen;
Modules of the block "STATISTICS" allow to realize full and detailed analysis of spectral-correlation
characteristics of geophisical fields by means of calculation of autocorrelation functions along each profile, crosscorrelation functions between separate profiles and different signs of the grid, two-dimensional autocorrelation and
cross-correlation functions as well as three-dimensional autocorrelation functions. There is possibility of the
estimation of one-dimension spectrums along separate profiles and two-dimension spectrums for coarse data.
Procedures of the calculation of statistical characteristics of geophisical fields in one- two- and three-dimensional
slithering windows allow getting fields of average, dispersion, asymmetry, kurtosis, mode, median and entropy of
the analyzed geophisical field.
In interpreting by fields of high order statistical moments the main interest presents the area of their extreme values
checking borders of geophisical field stationary areas. Their separation raises the efficiency of the solution of the
actual problem of geological zoning and mapping of any under investigation territory. Different fields of statistical
parameters emphasize different not obvious details of the source field, which can be effectively used in process of
the interpretation. It is possible to calculate statistical parameters of geophisical fields in dynamic windows that
automatically tuning on changing of spectral-correlation characteristics of the field along profiles, on area or in
space. This approach raises the accuracy of the estimation of the statistical characteristics of non-stationary
geophisical fields. It is become very important because practically all geophisical observations are not stationary.
In the block "STATISTICS" is included some modules for calculation of gradient characteristics of geophisical
fields (gradients of the field along profiles, between and layer of the grid, full gradient and its direction). Analysis
of gradient characteristics allows getting additional information about field’s structure and enhancing borders of the
stationary areas and anomalous areas. Except this, there is another one very interesting procedure in this section. It
is a module of the under investigation territory partition onto areas uniformed by field values and its gradient
characteristics.
Thus, the functional filling of this block helps to conduct the detailed analysis of geophisical fields by means of
study of their statistical and gradient characteristics, different correlation functions and spectrums.
4.1. Statistical characteristics...
Programs comprised into this block are intended for estimation of the statistical characteristics by field fragments or
in slithering windows with fixed or dynamic sizes and in windows of so called "alive form".
4.1.1. Statistical characteristics of field fragments
The program estimate the statistical characteristics of the sign of the grid or its fragment.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Copyright © 2003 MSGPU
Coscad 3Dt
Left picket
Right picket
Upper profile
Bottom profile
Upper layer
Bottom layer
37
–
–
–
–
–
–
left border of grid fragment by pickets;
right border of grid fragment by pickets;
upper border of grid fragment by profiles;
bottom border of grid fragment by profiles;
upper border of grid fragment by layers;
bottom border of grid fragment by layers.
The procedure estimates following statistical characteristics by any grid fragment:
1.
2.
3.
4.
1 n
Average of distribution m   xi ;
n i 1
n
1
2
Dispersion D 
 xi  m  ;

n  1 i 1
1 n
3
Asymmetry A    xi  m  D 3/ 2 ;
n i 1
Swing R  X max  X min of the correlation coefficient r0  m (m - minimum value of the argument where the
auto-correlation function is a zero R  m  0 );
5. Mode M 0  x _ f  x   max ;
6. Median Me_P(x < Me)= P(x > Me)= 0.5 (here f  x  , P  x  are a function of distribution density and a
function of the probability distribution of the random variate accordingly).
Information got by this module is used to choice an earl of further processing of geophisical field and to estimate
the stationarity of statistical and correlation characteristics of the field. Except this, by the value of the correlation
coefficient (radius of correlation) the width of the one-dimensional optimum filter is chosen.
Results of calculations are showed on the screen by means of module "Viewing the print file" from the section
“SERVICE”.
4.1.2. Statistical characteristics in the slithering window
This program calculates central statistical moments (the average, dispersion, asymmetry, kurtosis and standard) in
the two-dimensional slithering window of the fixed size.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Width of window (in pickets)
–
the width of the slithering window;
Height of window (in profiles)
–
the height of the slithering window;
Depth of window (in layers)
–
the depth of the slithering window;
First incline of window
–
slopping of three-dimensional window in planes of layers;
Second incline of window
–
slopping of three-dimensional window in planes of cuts;
Output grid
–
the result grid.
The program creates the grid containing five signs. The first sign is an average, the second is dispersion, the third is
an asymmetry, the fourth is a kurtosis and the fifth is an attitude of dispersion to average value (If in analyzed point
the average is a zero that attitude will be filled with dummy-code 131313E13. In this case use the program "Filling
of absent values" from the section “SERVICE”).
Analysis of values of the main statistical moments enables to get additional useful information about nature of the
distribution of sign’s values. Besides, (along with the other information) these distribution characteristics can be
effectively used in tasks of geological zoning by classifications programs of the complex.
Copyright © 2003 MSGPU
Coscad 3Dt
38
4.1.3. Statistical characteristics in the one-dimensional dynamic window
The program calculates six characteristics of the field of the source grid in the dynamic one-dimensional slithering
along profile window.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Width of base window (in pickets) –
Output grid
–
the source grid;
the processed sign of the source grid;
the base window width (!!! If zero - defined automatically);
the result grid.
The program creates the grid coinciding on size with the source grid and containing six signs:
1. Field of in-window average values;
2. Field of dispersion;
3. Field of asymmetry;
4. Field of kurtosis;
5. Factor of linear regression of the field in window;
6. Radius of correlation.
Enumerated characteristics are calculated in the slithering along profile window with variable sizes. Window width
changes depending on spectral-correlation characteristics of the field defined in the base window in the vicinity of
each point of the profile. The size of the base window must be not less then the power-hungriest forming of the
field on the profile. This parameter can be assigned by user for all profiles of the source grid or defined for each
separate profile automatically. For automatic choice of the window size it is enough to assign its equal zero.
Usage of the slithering window with variable size gives more proper estimation of the statistical characteristics of
the field in conditions when field is non-stationary along the profile. Analysis of values of statistical characteristics
allows to any researcher to get additional useful information about particularity of any geophisical field. So, in
fields of dispersions, asymmetries and kurtosis borders of anomalous object are more contrasting. By the regression
factor determines the slopping of graphics in the window. The radius of correlation defines the size of an anomaly
in concrete point. Besides, these distribution characteristics can be effectively used in tasks of geological zoning by
other programs of the complex.
4.1.4. Statistical characteristics in the two-dimensional dynamic window
This program is intended for calculation of first four statistical moments on the field of the source grid in the
dynamic two-dimensional slithering window.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Width of base window (in pickets) –
Height of base window (in profiles) –
Step by pickets
–
the source grid;
the processed sign of the source grid;
the base window width (!!! If zero - defined automatically);
the base window height (!!! If zero - defined automatically);
the parameter that allows reducing time of the calculations. If assign
value of this parameter equal 2 - time of the calculations decreases in abt
two times and etc.
Step by profiles (or by layers)
–
the parameter that allows reducing time of the calculations. If assign
value of this parameter equal 2 - time of the calculations decreases in abt
two times and etc.
Output grid
–
the result grid.
As a result of functioning the program is formed a grid containing following characteristics of the geophisical field
in each point of the grid:
1. Field of average values;
2. Field of dispersion;
3. Field of asymmetry;
4. Field of kurtosis.
Copyright © 2003 MSGPU
Coscad 3Dt
39
Enumerated characteristics are calculated in the slithering two-dimensional window with variable sizes. Window
width and height changes depending on spectral-correlation characteristics of the field defined in the base twodimensional window in the vicinity of each point of the profile. The size of the base window must be not less then
the power-hungriest forming of the field. This parameter can be assigned by user or defined for automatically. For
automatic choice of the window size it is enough to assign geometric parameters equal zero. Usage of the slithering
window with variable size gives more proper estimation of the statistical characteristics of the field in conditions
when field is non-stationary on area.
Analysis of values of statistical characteristics allows to any researcher to get additional useful information about
particularity of any geophisical field. So, in fields of dispersions, asymmetries and kurtosis borders of anomalous
object are more contrasting. By the regression factor determines the slopping of graphics in the window. The radius
of correlation defines the size of an anomaly in concrete point. Besides, these distribution characteristics can be
effectively used in tasks of geological zoning by other programs of the complex.
4.1.5. Statistical characteristics in the window of “alive form”
This program is intended for calculation of first four statistical moments on the field of the source grid in the twodimensional dynamic slithering window of “alive form”. The program computes the same characteristics as in the
module "Statistical characteristics in the two-dimensional dynamic window", but with greater reliability and
validity.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Width of base window (in pickets) – the base window width;
Height of base window (in profiles) – the base window height;
Output grid
–
the result grid;
Working mode
–
the window size is assigned by user or is chosen automatically;
Processing mode
–
processing switcher: by layers / by profiles.
As a result of functioning the program is formed a five-signed grid containing following characteristics of the
geophisical field in each point of the grid:
1. Field of average values;
2. Field of dispersion;
3. Field of asymmetry;
4. Field of kurtosis;
5. Field with the dispersion to average value ratio - a standard. (If in analyzed point the average is a zero that ratio
will be filled with dummy-code 131313E13. In this case use the program "Filling of absent values" from the
section “SERVICE”).
The usage of slithering window of the “alive form” allows getting the most reliable estimations of the statistical
moments of the field in conditions when field is non-stationary on area.
Analysis of values of statistical characteristics allows to any researcher to get additional useful information about
particularity of any geophisical field. So, in fields of dispersions, asymmetries and kurtosis borders of anomalous
object are more contrasting. The radius of correlation defines the size of an anomaly in concrete point. Besides,
these distribution characteristics can be effectively used in tasks of geological zoning by other programs of the
complex.
4.1.6. Estimation of the factor dispersion in the slithering window
This program is intended for calculation of factor and residual dispersion in the slithering window of the fixed
sizes.
Factor and residual dispersion is calculated using the following formulas:
D fact 
Copyright © 2003 MSGPU



N n
2
1 N
1
A

A
;
D

j

 Aij  A j
ost
N  1 j 1
N  n  1 j 1 i 1

2
Coscad 3Dt
40
Number of pickets in the window (window length) is associated with amount of factors, and number of profiles with amount of observation points for each factor. Also, the program returns factor-to-residual dispersions ratio in
each point of the grid.
F  D fact Dost
Using this field it is possible to reveal the areas, where there is influence of the factor (i.e. possibility of anomalies
presence).
Input parameters of the module:
Input grid
Sign of input grid
Width of the window (in pickets)
Height of the window (in profiles)
Window’s incline (in points)
Output grid
–
–
–
–
–
–
the source grid;
the processed sign of the source grid;
width of the window;
height of the window;
incline of the window;
the result grid.
As a result of functioning the program is formed a grid containing following four characteristics in each point of the
grid:
1. Factor-to-residual dispersions ratio F;
2. Factor dispersion Dfact;
3. Residual dispersion Dost;
4. Common (or total – a sum of Dfact and Dost) dispersion D.
4.2. Correlation characteristics...
In this block are united some modules allowing to estimate correlation characteristics of geophisical fields: one- and
two-dimensional autocorrelation functions ACF(m) and DACF(m,p), one- and two-dimensional cross-correlation
functions CCF(m) and DCCF(m,p), three-dimensional autocorrelation functions TACF(m,p,s) and etc.
4.2.1. Autocorrelation function
The program calculates an autocorrelation function for each profile of the source grid.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Maximum offset of ACF(m)
–
maximum of the argument value m (!!! If zero - defined automatically);
Output grid
–
the result grid.
This module calculates an autocorrelation function by the following formula:
1
R  m 
Nm
N m
f
i 1
i

 f average   fi  m  f average

By the autocorrelation function you can estimate a linear size of anomaly on profile and choose the parameters of
the one-dimensional optimum filter.
The program creates a grid containing values of the autocorrelation function for the argument from zero up to
maximum offset. Since the autocorrelation function is even that the function calculates just for positive values of
the argument.
Amount of profiles and layers in the resulting grid complies with amount of profiles and layers in the source grid,
but amount of pickets complies with amount of offsets m for which the autocorrelation function R(m) is calculated.
The value of the autocorrelation function is corresponding to zero value of the argument. Viewing the
autocorrelation function by means of graphic modules allows to select the area of non-stationary correlation
characteristics of the field on area. By the radius of correlation the depth of anomaly objects.
Copyright © 2003 MSGPU
Coscad 3Dt
41
4.2.2. Cross-correlation function between profiles
This program is calculates a cross-correlation function CCF(m) between neighboring profiles of the source grid
layers.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Maximum offset of CCF(m)
–
the source grid;
the processed sign of the source grid;
maximum of the argument value m of CCF(m) by pickets
(!!! If zero - defined automatically);
Output grid
–
the result grid;
Processing mode
–
processing switcher: between neighbor profiles of layers / of cuts.
Usage of this module allows to estimate the correlation relationship between observations on two neighbor profiles
to a grid ( f 1 and f 2 ) by means of estimation of their cross-correlation function:
B  m 
1
Nm
N m
f
i 1
1
i

1
2
 f average
  fi2 m  faverage

Using the cross-correlation function it is possible to emphasize areas of the stationarity breach of correlation
characteristic of the field, which are often connected with tectonic breaches. On values of the argument mext
corresponding to extreme values B  mext  it is possible to estimate the correlation directions of the field from
profile to profile and energy of signal-to-noise merit   B  metr  1  B  metr   . It is also provided estimation of
the cross-correlation function between two signs of the grid. The analysis of the cross-correlation function in this
case allows to separate the areas with presence and absences of correlation between two different signs.
The program creates a grid containing values of the cross-correlation function. Amount of profiles in the resulting
grid is less than in the source by one. The first profile of the output grid contains values of cross-correlation
function between the first and the second profile of input grid, the second profile contains values of crosscorrelation function between the second and the third profile and etc. Values of the cross-correlation function are
written into the output grid in following order: B(-max), B(-max+1), …, B(0), B(1), …, B(max-1), B(max), where
max is the maximum of the argument value m of CCF(m) determined by user.
4.2.3. Cross-correlation function between fields
This program calculates a cross-correlation function between two different signs of two grids B(m).
Input parameters of the module:
Input grid
Sign of input grid
Second input grid
Sign of second input grid
Maximum offset of CCF(m)
–
–
–
–
–
the source grid;
the processed sign of the source grid;
the second source grid;
the processed sign of the second source grid;
maximum of the argument value m of CCF(m) by pickets
(!!! If zero - defined automatically);
Output grid
–
the result grid;
After functioning the program creates the grid containing values of cross-correlation function. Each profile of the
resulting grid contains cross-correlation function between two signs on each of profiles.
Values of the cross-correlation function between signs (by each profile) are written into the output grid in following
order: B(-max), B(-max+1), …, B(0), B(1), …, B(max-1), B(max), where max is the maximum of the argument
value m of CCF(m) determined by user. The analysis of the cross-correlation function allows separating the areas
with presence and absences of correlation between two different signs. Except this, the big interest presents the
value of the offset under which the correlation between signs on the separate profile is existed.
Copyright © 2003 MSGPU
Coscad 3Dt
42
4.2.4. Two-dimensional autocorrelation function
The module estimates correlation characteristic of the field on area by calculation of a two-dimensional
autocorrelation function.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Maximum offset of DACF(m,p) by pickets – maximum step m between pickets for the two-dimensional
autocorrelation function DACF(m,p) (!!! If zero - defined automatically);
Maximum offset of DACF(m,p) by profiles – maximum step p between profiles for the two-dimensional
autocorrelation function DACF(m,p) (!!! If zero - defined automatically);
Output grid
–
the result grid;
Processing mode
–
processing switcher: calculate by layers / by cuts.
This module calculates an autocorrelation function by the following formula:
D  m, p  
1
1
M p Nm
M p Nm
 f
j 1
i 1
i, j

 f average   fi  m , j  p  f average

By the two-dimensional autocorrelation function the correlation directions of the field is estimated. These
directions usually comply with the spread of the most power-hungry anomaly. Analysis of the structure of the twodimensional autocorrelation function allows to do a motivated choice of the size and slopping of window to
filtering for two-dimensional linear filters. The program allows to calculate the two-dimensional autocorrelation
function DACF(m,p) for signs of the grid (fully or by the cut).
The two-dimensional autocorrelation function is the most important correlation function describing correlation
characteristics of the field on area. By the two-dimensional autocorrelation function it is possibly to judge about
presented correlation directions, which comply with spreads of the most power-hungry anomaly.
4.2.5. Two-dimensional cross-correlation function
This module calculations the two-dimensional cross-correlation function between two f 1 and f 2 signs of the grid:
1
1
DCCF  m, p  
M p Nm
M p Nm
 f
j 1
i 1
1
i, j
1
2
 f average
   fi2m, j  p  faverage

Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Maximum offset of DCCF(m,p) by pickets – maximum step m between pickets for the two-dimensional crosscorrelation function DCCF(m,p) (!!! If zero - defined automatically);
Maximum offset of DCCF(m,p) by profiles – maximum step p between profiles for the two-dimensional crosscorrelation function DCCF(m,p) (!!! If zero - defined automatically);
Output grid
–
the result grid;
Processing mode
–
processing switcher: calculate between layers / between cuts.
By means of the two-dimensional cross-correlation function are estimated the correlation relationship between two
signs on the area.
4.2.6. Three-dimensional autocorrelation function
This program is intended for calculation of a three-dimensional autocorrelation function TACF(m,p,k) for a sign of
a grid.
Input parameters of the module:
Input grid
–
Copyright © 2003 MSGPU
the source grid;
Coscad 3Dt
43
Sign of input grid
–
the processed sign of the source grid;
Maximum offset of TACF(m,p,k) by pickets –
maximum step m between pickets for the three-dimensional
autocorrelation function TACF(m,p,k)
(!!! If zero - defined automatically);
Maximum offset of TACF(m,p,k) by profiles –
maximum step p between profiles for the three-dimensional
autocorrelation function TACF(m,p,k)
(!!! If zero - defined automatically);
Maximum offset of TACF(m,p,k) by layers –
maximum step k between layers for the three-dimensional
autocorrelation function TACF(m,p,k)
(!!! If zero - defined automatically);
Output grid
–
the result grid;
The three-dimensional autocorrelation function allows to study the correlation characteristics of the field in space.
On the base of TACF(m,p,k) the calculation of weighting coefficients of three-dimensional filters is realized.
4.2.7. Correlation coefficient in the slithering window
The Program calculations the usual correlation coefficient and the rank correlation coefficient of Spirmen between
two signs in a slithering three-dimensional window with the fixed size. The window height (amount of profiles or
layers) can be is 1 (to work with two-dimensional grids).
Input parameters of the module:
Input grid
Sign of input grid
Second input grid
Sign of second input grid
Window width
Window height
–
–
–
–
–
–
the source grid;
the processed sign of the source grid;
the second source grid;
the processed sign of the second source grid;
the window width in pickets for correlation coefficient counting;
the window height in profiles (or layers) for correlation coefficient
counting;
Step by pickets
–
the parameter that allows reducing time of the calculations. If assign
value of this parameter equal 2 - time of the calculations decreases in abt
two times and etc.
Step by profiles (or by layers)
–
the parameter that allows reducing time of the calculations. If assign
value of this parameter equal 2 - time of the calculations decreases in abt
two times and etc.
Output grid
–
the result grid.
Working mode
–
you may calculate the usual correlation coefficient and the rank
correlation coefficient of Spirmen;
Processing mode
–
processing switcher: by layers / by cuts.
Module allows to estimate the value of the usual correlation coefficient:
rxy 
1
n   x y
N
 x  x  y  y 
i
i 1
i
( x , y ,  x y are estimations of average values and mean square deviation of signs X and Y accordingly) and the
rank correlation coefficient of Spirmen:
6  ri x  ri y 
n
 xy  1 
i 1
n
3
 n
( ri x , ri y are ranks corresponding to sign values X and Y) between two signs in the slithering two-dimensional
window of the fixed size.
Analysis of values of correlation coefficient on area enables to select the areas of presence or absences of
correlation between signs that can be effectively used in problem of geological zoning and mapping.
Copyright © 2003 MSGPU
Coscad 3Dt
44
4.3. Spectral characteristics...
Spectral analysis occupies the central place in processing of geophisical data. The decomposition of the observed
field onto different frequency components already gives much information about structure of the field. Herewith, it
is important to emphasize the applicability of spectral analysis for description characteristics of geophisical fields
given as deterministic or casual functions. Spectral analysis has very broad and varied possibilities in filtering of
raw data, estimation of inaccuracy and comparison of data processing efficiency. This section contains the modules
of calculation of the univariate Fourier spectrum and two-dimensional quick Fourier transformation.
4.3.1. Univariate spectrum
The program calculates the Fourier spectrum for each profile of the grid.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Output grid
–
the source grid;
the processed sign of the source grid;
the result grid.
By means of this procedures it is possible to calculate the amplitude
Rm  arctg  Bm Am  , m  0,1,
Am 
1
N
N
1
 2 mi 
 , Am 
N 
N
 f cos 
i 1
i
,N /2
Fourier
N
spectrums
on
all
Rm  Am2  Bm2
profiles
of
the
and phase
grid
(where
 2 mi 
 and m is a harmonica number, N is an amount of pickets
N 
 f sin 
i 1
i
of the grid profile).
The program creates the two-signs grid: energy spectrum for frequencies from zero up to Naykvist frequency and
phase spectrum for the same frequencies.
Amount of profiles and layers in the resulting grid complies with amount of profiles and layers in the source grid,
but amount of pickets is a half plus one concerning to amount of pickets in the source grid plus one. This complies
with amount of harmonicas in a spectrum. The first picket in the resulting grid corresponds to the first (zero)
spectrum harmonica and etc.
4.3.2. Two-dimension spectrum
This program realizes the algorithm of quick two-dimensional Fourier transformation in suggestion that number of
pickets and profiles in the grid is multiple to degree of numbers two, three, five or seven.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Output grid
–
the result grid;
Processing mode
–
processing switcher: by layers / by cuts.
If sizes of the grid do not satisfy to the putted condition than they will be automatically cut with minimum losses of
information for estimation of the two-dimensional spectrum.
The program creates the grid containing three signs: energy, phase and amplitude spectrums. The first picket of the
first profile of the resulting grid contains values of the spectrum corresponding to zero frequency on pickets and
zero frequency on profiles (layers). The last picket of the last profile in the resulting grid contains values of the
spectrum corresponding to Naykvist frequency on pickets and profiles (layers).
Analysis of the two-dimensional spectrum allows to estimate the spectral composition of the field, amplitudes
composing its harmonicas and anomaly parameters presented by low-frequency part of spectrum and noise
presented by high-frequency part of spectrum.
Copyright © 2003 MSGPU
Coscad 3Dt
45
4.4. Gradient characteristics
The given program calculates of the field gradient along profiles or transversely profiles spread, between layers of
the grid, full gradient of the field and directions of the full gradient in plane of the layers or in cut’s plane. The
direction of the full gradient calculates in radians.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Output grid
–
the source grid;
the processed sign of the source grid;
the result grid.
When the two-dimensional grid is processing the four-signs grid is formed:
First sign
–
a gradient of the field along profiles.
Second sign
–
a gradient of the field between profiles (layers).
Third sign
–
a full gradient in a plane of profiles (layers) and pickets.
Fourth sign
–
a direction of the full gradient
When the three-dimensional grid is processing the grid with eight signs is formed:
First sign
–
a gradient of the field along profiles.
Second sign
–
a gradient of the field between profiles.
Third sign
–
a full gradient in a plane of profiles and pickets.
Fourth sign
–
a direction of the full gradient in a plane of profiles and pickets in degrees. The direction N-S
corresponds to the 90 degrees angle and direction E-W corresponds to the angles 0 and 180
degrees.
Fifth sign
–
a gradient of the field between layers.
Sixth sign
–
a full gradient in a plane of layers and pickets.
Seventh sign –
a direction of the full gradient in a plane of layers and pickets in degrees.
Eighth sign
–
a direction of the full gradient in the space XYZ.
Analysis of gradient characteristics of the field enables to get for researcher additional useful information about
particularity of a geophisical field. The borders of anomalous object on corresponding directions are risen above in
gradient fields. When interpreting it is necessary to account following:
 Borders of anomalous objects are noted by extremums in the field gradient along axes and by maximums in the
fields of the full gradient;
 By extremums in fields of gradient characteristics borders of anomalies with different amplitudes are noted,
that allows to see simultaneously sidebars anomaly of different amplitudes during the visualization;
 Gradient characteristics along a certain direction allow to emphasize the borders of anomaly with perpendicular
spread for this direction;
 The field of the full gradient direction allows to estimate anomaly spread in each point of the source grid of the
observations, but contrasting transition from minimum values to maximum checks the position of the
anomaly’s axes.
Along with the other information these characteristics of the field can be used in tasks of geological zoning
taxonomic programs of the complex. Most effectively, by means of gradient characteristics, dares tasks of the
highlighting of the borders between anomalies or stationary areas.
4.5. Sounding...
In this section are presented programs realizing original approach to estimation the changes in statistical and
correlation characteristics (with depth) of the field on the base of their calculation in slithering windows with
different sizes. The possibility of using of probabilistic-statistical methods at building of physico-geological models
and estimation of geometric parameters anomalies object are considered.
Copyright © 2003 MSGPU
Coscad 3Dt
46
4.5.1. Statistical sounding
This program calculates first four central statistical moments in the window with different sizes with forming of the
three-dimensional grids. Herewith, the first layer to the resulting grid is the result of moment’s calculation in the
window with minimum size and the last layer - with maximum size of the window.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Start window size
–
the source grid;
the processed sign of the source grid;
the initial size of the base slithering window in which the radius to
correlation is estimated;
(!!! If zero - defined as 3)
Final window size
–
the final size of the base slithering window in which the radius to
correlation is estimated;
(!!! If zero - defined as 0.5*MIN[amount of pickets or profiles]);
Step of detailing in points
–
the step of the window sizes changes (from initial to final size).
Than the value of this parameter bigger than accounting time is less.
Herewith the quality of the final result is falling.
Output grid
–
the result grid.
After calculations the three-dimensional grid coinciding on pickets and profiles with the source one is being
formed. Amount of layers of the resulting grid is defined by expression: [final – initial window sizes]/2+1.
Amount of signs in the resulting grid is 4. The first sign contains the average value in the window, the second – the
dispersion, the third – the asymmetry and the fourth – the kurtosis. Each layer of the resulting grid contains values
of the statistical moments calculated in windows with corresponding sizes. The first layer of the resulting grid
corresponds to the window with "initial size", the last - "final size". The proposed statistical sounding allows to
track the changes in statistical characteristics computed in slithering windows depending on the analyzed forming
frequencies of the field. Thus, the first layer of will present values of the statistical moments of high-frequency
forming of the field, the last one - an low-frequency forming. Considering the fact that extreme values of the field
of dispersions, asymmetries and kurtosis is checking areas of stationary breaches, it is possible to track the position
of these borders for different (on size and locations) geological objects.
4.5.2. Correlation sounding
This program calculates the two-dimensional correlation radius of areal data.
The correlation radius is written into the resulting grid with minus sign. For example the value r=-5090.3 follows to
interpret as r=5090.3. The minus sign is putted because of orientation of the Z-axis. Thereby, for example, values
r=-100.2 and r=-320.1 are indicative that in the second event anomaly source is situated deeper then in the first one.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Start window size
–
the source grid;
the processed sign of the source grid;
the initial size of the base slithering window in which the radius to
correlation is estimated;
(!!! If zero - defined as 3)
Final window size
–
the final size of the base slithering window in which the radius to
correlation is estimated;
(!!! If zero - defined as 0.5*MIN[amount of pickets or profiles]);
Step of detailing in points
–
the step of the window sizes changes (from initial to final size).
Than the value of this parameter bigger than accounting time is less.
Herewith the quality of the final result is falling.
Output grid
–
the result grid.
After calculations the three-dimensional grid coinciding on pickets and profiles with the source one is being
formed. Amount of layers of the resulting grid is defined by expression: [final – initial window sizes]/2+1.
The each layer of the resulting grid contains the correlation radius value calculated in the window with
corresponding sizes. The first layer of the resulting grid corresponds to the window with "initial size" and the last Copyright © 2003 MSGPU
Coscad 3Dt
47
"final size". The more size of the window the more depth of the sounding of anomalous sources. Thereby, assigning
different values of the interval of the window’s size changing you are sounding the position of anomalous sources
on the determined interval of the depths. It is recommended the joint interpretation of results got by means of the
given module and by the program "Statistical estimation of anomalous object’s parameters".
The question of the two-dimensional correlation radius determination is very actual because it allows to estimate
the depth of the sources anomaly location by formulas gotten by S.A. SERKEROV for data of gravitational and
magnetic explorations. For the estimation of the true depth of the sources multiply the location the field of the
correlation radius on the corresponding factor (see S.A. SERKEROV) having used by program "Some
transformations of data " from the section "SERVICE".
Working with the program it is recommended to execute the following actions:
1. Using the procedure "Interpolation of grids" from the section "SERVICE" bring the source grid to uniform
(dx=dy) because the procedure works just with uniform grids;
2. For increasing of the program efficiency extrapolate the grid with usage of the module "Extrapolation of grids"
with zero value of amount points of extrapolations beforehand. Herewith the source grid will be added on the
left and on the right by additional pickets and from above and from below by additional profiles. The amount of
added points will be approximate 0.1 from the amount of pickets and profiles;
3. The maximum window’s size for processing of the extrapolated grid must be not greater than
0.5*MIN[amount of pickets or of profiles] of the source (before extrapolation) grid;
4. After processing of the extrapolated grid by this program (for reduction of the resulting grid to the source size)
use the module "Fragmentation of grids " having sliced extrapolation points;
5. The input parameter "Step of detailing in points" allows to reduce time of the calculations. Than the value of
this parameter bigger than less the accounting time. Herewith the quality of the final result is falling.
4.5.3. Cross-correlation sounding
This program calculates the correlation coefficient in the slithering window of the different size and forms the
three-dimensional grid. Herewith, the first layer of the resulting grid is the calculation result of the correlation
coefficient in the window with minimum size, the last layer - with maximum window size.
Input parameters of the module:
Input grid
Sign of input grid
Second input grid
Sign of second input grid
Start window size
–
–
–
–
–
the source grid;
the processed sign of the source grid;
the second source grid;
the processed sign of the second source grid;
the initial size of the base slithering window in which the radius to
correlation is estimated;
(!!! If zero - defined as 3)
Final window size
–
the final size of the base slithering window in which the radius to
correlation is estimated;
(!!! If zero - defined as 0.5*MIN[amount of pickets or profiles]);
Step of detailing in points
–
the step of the window sizes changes (from initial to final size).
Than the value of this parameter bigger than accounting time is less.
Herewith the quality of the final result is falling.
Output grid
–
the result grid.
After calculations the three-dimensional grid coinciding on pickets and profiles with the source one is being
formed. Amount of layers of the resulting grid is defined by expression: [final – initial window sizes]/2+1.
The each layer of the resulting grid contains the correlation coefficient calculated in the window with
corresponding sizes. The first layer of the resulting grid corresponds to the window with "initial size" and the last "final size". The step value is 1 by default.
The proposed cross-correlation sounding allows to track the correlation relationship changes depending on
analyzed frequency forming of the source field. The first layer of the resulting grid will characterize the correlation
relationship between fields high-frequency forming, the last - an low-frequency forming.
Copyright © 2003 MSGPU
Coscad 3Dt
48
4.5.4. Gradient sounding
This program calculates gradient characteristics with consequent discharging of the source grid. Such procedure
allows to research changes of gradient characteristics with depth. Herewith, the first layer of the resulting grid is a
result of the gradient characteristics calculation with minimum source grid discharging, the last layer - with
maximum discharging.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Start grid discharging
–
the source grid;
the processed sign of the source grid;
the initial discharging in points;
(!!! If zero - defined as 1)
Final grid discharging
–
the final discharging in points;
(!!! If zero - defined as 0.25*MIN[amount of pickets or profiles]);
Step of discharging in points
–
the step of the discharging changes (from initial to final).
Than the value of this parameter bigger than accounting time is less.
Herewith the quality of the final result is falling.
Output grid
–
the result grid.
After calculations the three-dimensional grid coinciding on pickets and profiles with the source one is being
formed. Amount of layers of the resulting grid is defined by expression: [final – initial window sizes]/2+1.
After processing the four-signs grid is formed:
First sign
–
a gradient of the field along profiles.
Second sign
–
a gradient of the field between profiles (layers).
Third sign
–
a full gradient in a plane of profiles (layers) and pickets.
Fourth sign
–
a direction of the full gradient
4.6. Estimation of anomalous object’s parameters.
This section contains modules allowing to estimate anomalous object’s parameters by means of usage probabilisticstatistical methods of the data analysis.
4.6.1. Estimation by I.I. Priezzhev
The program calculates the equivalent sharing the masses of the sources at the depth.
Input parameters of the module:
Input grid
Sign of input grid
Layer of input grid
Angle of T-vector with horizon
Angle of J-vector with horizon
Output grid
Field type
–
–
–
–
–
–
–
the source grid;
the processed sign of the source grid;
the processed layer of the source grid;
the angle of the full vector of the magnetic field T with horizon;
the angle of the vector of the magnetizing J with horizon;
the result grid;
it is possible to solve the inverse problem for vertical forming magnetic
field dZ, gravitational field or for the full vector of the magnetic field dT.
After functioning the program forms the grid with N pickets, M profiles and N/3 layers (where N is amount of
pickets and M is amount of the profiles of the source grid).
In each point of the gotten cube the relative density of the sources magnetic or gravitational field is kept. The
distance between layer of the resulting grid is a distance between pickets of the source grid that allows to estimate
the position of outraging solids.
Copyright © 2003 MSGPU
Coscad 3Dt
49
4.6.2. Estimations by A.V. Petrov
This program is intended for estimation of the geometrical and relative sharing of masses of the anomalous sources.
The program is based on integrated usage of statistical, spectral-correlation methods and the algorithm of adaptive
filtering in the window of the “alive form”.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Start window size
–
the source grid;
the processed sign of the source grid;
the initial size of the base slithering window in which the radius to
correlation is estimated;
(!!! If zero - defined as 3)
Final window size
–
the final size of the base slithering window in which the radius to
correlation is estimated;
(!!! If zero - defined as 0.5*MIN[amount of pickets or profiles]);
Step of detailing in points
–
the step of the window sizes changes (from initial to final size).
Than the value of this parameter bigger than accounting time is less.
Herewith the quality of the final result is falling.
Output grid
–
the result grid;
Method
–
it is possible to solve the problem using the adaptive energy filtering,
entropic filtering, median filtering and simple averaging in a window.
As a result of calculations the three-dimensional grid is formed. This grid contains 3 signs and coinciding on
pickets and profile with the source one. Amount of layers of the resulting grid is defined by expression: [final –
initial window sizes]/2+1.
First sign
–
an equivalent masses sharing.
Second sign
–
masses sharing located on small depth (similarly to downward recalculation).
Third sign
–
an analogue of the upward recalculation.
It is recommended the joint interpretation of results gotten by means of the given module and programs
"Correlation sounding", "Statistical sounding" and "Cross-correlation sounding".
4.6.3. Tracing of axes of anomaly
This program is intended for tracing of anomaly with different energy from profile to profile. For tracing the
original modification of the univariate adaptive filtering is used.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Minimal anomaly size
–
the size in points of the most short selected anomalies;
Threshold in % from mean square deviation
–
a parameter allowing to adjust the sensitivity of the algorithm.
The less value of the parameter the bigger the sensitivity to weak
anomaly.
Output grid
–
the result grid;
Working mode
–
a size of the selected anomaly can be assigned by user or chosen
automatically.
As a result of functioning the program is formed the grid containing two signs. The first sign contains the statistics
with maximums correspond to the axes of positive anomalies and minimums - axes of negative anomalies. Second
sign contains parameter which can takes three values:
1 – the positive anomaly is present;
-1 – the negative anomaly is present;
0 – Nothing is present.
Working with the program it is necessary to bear in mind that well tracing anomaly axes located across to the
profile’s spread. For highlighting of anomaly spreads with the same direction as the profile’s spread it is
recommended to turn the grid on 90 degrees left to right, or right to left using the module "Rotation of grids".
Copyright © 2003 MSGPU
Coscad 3Dt
50
4.6.4. Depth estimation of the main anomalous surfaces
This program is intended for evaluation of depths of the main gravel-magnetic surfaces on the base of the analysis of
the logarithm of spectrum in slithering window and field’s correlation radius.
Input parameters of the module:
Input grid
Sign of input grid
Layer of input grid
Radius of the window
–
–
–
–
the source grid;
the processed sign of the source grid;
the processed layer of the source grid;
the most important parameter of the algorithm, assigned in the grid’s
points. It has to be assigned in 3-5 times greater than the expected depth
of the main gravel-magnetic surfaces;
Step of detalization
–
the step (in points of the grid) in which the estimation of the depth will be
realized. It is recommended to conduct the estimation in each fifth point
of the source grid;
Working mode
–
with trend removing and without removing;
Output grid
–
the result grid.
It is recommended to choose the mode with trend removing.
As a result of functioning the program is formed a grid, coinciding on size with the source one and containing 3 signs.
The first sign contains the evaluation of an upper edge of the main gravel-magnetic surface on the base of analysis of
the logarithm of the spectrum. The second one contains a lower edge. The last one contains the estimation of a lower
edge of main gravel-magnetic surface on the base of analysis of the correlation radius.
Copyright © 2003 MSGPU
Coscad 3Dt
51
5. FILTERING
Trend-analysis and filtering of geophisical fields occupy the most important place in the processing of geologygeophisical data. Program modules realizing this stage of the processing are united in the block "FILTERING" of
the computer technology "COSCAD3Dt". They allow to solve the following problems of the processing and
interpretation of geology-geophisical observations:
 Separations of fields on components with estimation of the form and parameter separated component;
 Estimations of the form of low-frequency trend forming of geophisical fields;
 Separations and estimations of the form of weak anomalies with the amplitude commensurable or less than noise
level.
This block includes the original realizations of one-, two- and three-dimensional Viner’s, matched and energy
optimum filtering, algorithms of entropy filtering solving a problem of the efficient processing of (geochemical,
radiometric, ecological) geophisical fields with presence of hurricane values.
In this block different modifications of one- and two-dimensional polynomial filters are broadly presented. The
each module in the given block was developed with provision of following principle positions:
 Abilities of the correct functioning of the algorithm in conditions of the absence to a priori information on
spectral-correlation characteristics of useful signals and noise that is corresponds to the most often existing
situation when the real geology-geophisical observations are processing;
 Possibilities of the processing of very-large arrays of input information within the frames of real time
intervals;
 The maximum account in the algorithm of non-execution of the requirement of data stationarity;
 Realizations of the program of the concrete filter in time area for the reason of exceptions of disadvantage
effects connected with distortion of useful signal and noise parameters when turning into and out of the
spectral area using the Fourier transformation.
5.1. One-dimension filtering...
Different realizations of univariate filters are presented in this section.
5.1.1. One-dimension filtering in the fixed window
This program is intended for on-profile filtering of data in the grid.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Window size (in pickets)
–
the size of filtration window (!!! If zero - defined automatically);
Output grid
–
the result grid;
Working mode
–
the base window size can be assigned by user or chosen automatically;
In this program there are several on-profile filters:
Univariate energy filtering
–
this filter occupies the intermediate position between the KolmogorovViner’s filter and the detection filter. The criterion of an optimal signal on
output of this filter is a maximization of energy signal-to-noise merit.
Univariate entropy filtering
–
allows to process fields with presence of hurricane values (geochemical,
radiometric etc.)
Univariate median filtering
–
calculates the median in slithering window.
Univariate correlation filtering –
is useful for estimation of the noise in the small-size window and least of
all changes the data.
Univariate trend filtering
–
efficient in problems of trend removing.
Averaging in the univariate window – calculates average in the slithering window.
Copyright © 2003 MSGPU
Coscad 3Dt
52
After functioning the program creates the grid containing two signs: 1 st is local (remaining) and 2nd - regional
(trend) forming of the source grid.
5.1.2. One-dimensional polynomial filtering
This program is intended for on-profile filtering of separate signs of the grid using approximation of field values on
profile with polynomials (degrees from 1 up to 49).
Input parameters of the module:
Input grid
Sign of input grid
Window size (in pickets)
Degree of polynomial P
Output grid
Working mode
–
–
–
–
–
–
the source grid;
the processed sign of the source grid;
the size of filtration window (!!! If zero - defined automatically);
the degree of approximating polynomial P<50;
the result grid;
the base window size can be assigned by user, taken equal to profile
length or chosen automatically.
There is a possibility to conduct filtering in the window mode (in slithering along profiles window) and in mode of
the building approximating polynomial on all point of the profile. In the last case the window width is assigned
equal to the profile length in pickets. Usage of this program is effectively in processing of any types of geophisical
fields.
The dignity of this algorithm is a possibility to change the frequency characteristics of the signal on outputting of
the filter and to change the degree of approximating polynomial. So, the less the polynomial degree the less of
high-frequency components remains in the output signal of this filter.
The program creates the grid containing two signs: local (remaining) and regional components of the source field.
5.1.3. One-dimensional adaptive filtration
This program is intended for on-profile filtering of geophisical fields by an adaptive filter. The univariate adaptive
filter allows effectively processing the non-stationary along profile fields.
Input parameters of the module:
Input grid
Sign of input grid
Size of base window (in pickets)
Step by pickets
–
–
–
–
the source grid;
the processed sign of the source grid;
the size of the base filtration window (!!! If zero - defined automatically);
the parameter that allows reducing time of the calculations. If assign
value of this parameter equal 2 - time of the calculations decreases in abt
two times and etc.
Output grid
–
the result grid;
Filter type
–
there are several filters in this program: correlation, energy, entropy,
median filtering and simple time averaging.
Working mode
–
the base window size can be assigned by the user or chosen
automatically;
In contrast to usual filters, which do not change their own parameters during the filtering process, the given filter
automatically changes values of weighting coefficients and size of the filtering window depending on changes of
spectral-correlation characteristics in a vicinity of the base window in each points of the field. This allows to
process the non-stationary on spectral-correlation characteristics geophisical fields correctly.
The size of the base window must be odd and not more than the lengths of the most power-hungry component of
the field on profile and not less than three. This parameter can be assigned by user or be chosen automatically if
assign its value equal zero.
The program creates the grid containing two signs: 1st sign is a local (remaining) component and 2nd sign is a
regional (the most power-hungry) component of the given field.
Copyright © 2003 MSGPU
Coscad 3Dt
53
5.2. Two-dimension filtering...
Different realizations of two-dimensional filters are presented in this section.
5.2.1. Two-dimension filtering in the fixed window
This program is intended for two-dimensional filtering of the field for division reasons onto regional and local
components by using of two-dimensional energy and entropy filtering.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Window width (in pickets)
–
Window height (in profiles/layers) –
Incline of filtration window
–
Output grid
–
Filtration type
–
the source grid;
the processed sign of the source grid;
the width of the filtration window (!!! If zero - defined automatically);
the height of the filtration window (!!! If zero - defined automatically);
the slopping of the filtration window;
the result grid;
energy filter, which is maximize the energy signal-to-noise merit on the
output. The filter occupies intermediate position between the
Kolmogorov-Viner’s filter and the detection filter;
entropy filter, which is intended for filtering of fields with presence of
hurricane values (geochemical, radiometric etc.);
Processing mode
–
two-dimensional filtering can be provided by layers and by cuts of the
source grid.
The filtering window sizes are chosen on the base of the analysis of two-dimensional autocorrelation function of
the source grid or are assigned coming from nature of the problem. Obviously that the more the window size the
less the share of the high frequency in the trend component.
The program creates the grid containing two signs: the local (remaining) and the regional components of the source
field.
5.2.2. Two-dimensional polynomial filtering
This program is intended for division of the field onto the regional and the local components by building of the
two-dimensional polynomial (from zero up to ninth degree inclusive). This filtering is realized in two-dimensional
slithering window with the fixed size. The window can move along the profile or by profiles with bigger than unit
step that allows to enlarge the velocity of the performing the program.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Window width (in pickets)
–
Window height (in profiles/layers) –
Step by pickets
–
Step by profiles (or by layers)
–
Output grid
Working mode
–
–
Processing mode
–
Copyright © 2003 MSGPU
the source grid;
the processed sign of the source grid;
the width of the filtration window (!!! If zero - defined automatically);
the height of the filtration window (!!! If zero - defined automatically);
the parameter that allows reducing time of the calculations. If assign
value of this parameter equal 2 - time of the calculations decreases in abt
two times and etc.
the parameter that allows reducing time of the calculations. If assign
value of this parameter equal 2 - time of the calculations decreases in abt
two times and etc.
the result grid;
the base window size can be assigned by user, chosen automatically or
taken equal to profile length. In the last case the window width is
assigned equal to the profile length in pickets.
two-dimensional filtering can be provided by layers and by cuts of the
source grid.
Coscad 3Dt
54
The filtering window sizes are chosen on the base of the analysis of two-dimensional autocorrelation function of
the source grid or are assigned coming from nature of the problem. Obviously that the more the window size the
less the share of the high frequency in the trend component.
Usage of this program is effectively in processing of any types of geophisical fields.
The dignity of this algorithm is a possibility to change the frequency characteristics of the signal on outputting of
the filter and to change the degree of approximating polynomial. So, the less the polynomial degree the less of
high-frequency components remains in the output signal of this filter. Except this the automatic orientation of
approximating polynomial onto the correlation direction of the field in window, draws near this filter to adaptive in
processing of non-stationary on area geophisical fields is processing.
The program creates the grid containing two signs: the local (remaining) and the regional components of the source
field.
5.2.3. Two-dimension adaptive filtration
The program is intended for two-dimensional adaptive filtering of geophisical fields.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Window width (in pickets)
–
Window height (in profiles/layers) –
Step by pickets
–
the source grid;
the processed sign of the source grid;
the width of the filtration window (!!! If zero - defined automatically);
the height of the filtration window (!!! If zero - defined automatically);
the parameter that allows reducing time of the calculations. If assign
value of this parameter equal 2 - time of the calculations decreases in abt
two times and etc.
Step by profiles (or by layers)
–
the parameter that allows reducing time of the calculations. If assign
value of this parameter equal 2 - time of the calculations decreases in abt
two times and etc.
Output grid
–
the result grid;
Filter type
–
there are several filters in this program: energy, entropy, median filtering
and simple time averaging.
Working mode
–
the base window size can be assigned by the user or chosen
automatically;
Processing mode
–
two-dimensional filtering can be provided by layers and by cuts of the
source grid.
Essentially the two-dimensional adaptive filtering of geophisical fields is reduced to automatic changing of filter
parameters (width, height, slopping filtering window and weighting coefficient) when some spectral-correlation
characteristics of the field is changing in vicinity of the base window for each point of the field.
Base window sizes are assigned by a user and must be not less than sizes of the most power-hungry component of
the field. Usually the size is chosen on the base of analysis of two-dimensional autocorrelation function of the
source grid. The usage of this two-dimensional adaptive filter allows effectively processing the non-stationary on
area (by spectral-correlation characteristics) fields. Since majority of real observed fields are non-stationary on area
the usage of this program gives the more correct results than usage of usual two-dimensional filters with constant
parameters.
In the sense of optimum criterion of output signal in this module several filters are marketed: energy filter, which is
maximize the energy signal-to-noise merit; entropy filter - intended for filtering of non-potential (the radiometric,
geochemical fields with presence of hurricane values) and simple time averaging in window.
The program creates the grid containing four signs: the local (remaining) field, the regional field (trend), the slop of
filtering window (i.e. the direction of correlation of the field in the point) and the radius to correlation for
estimation of occurrence depth of anomalous objects. The direction is defined by offset (expressed in pickets) of
correlation directions on the following profile relatively to previous. The positive direction corresponds to increase
of the sequence number of pickets, negative – to their reduction.
Copyright © 2003 MSGPU
Coscad 3Dt
55
5.2.4. Two-dimension filtering in the window of “alive form”
It is the BEST program for two-dimensional filtering of non-stationary geophisical fields.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Base window width (in pickets) –
the width of the base filtration window;
Base window height (in profiles/layers) – the height of the base filtration window;
Output grid
–
the result grid;
Filter type
–
there are several filters in this program: energy, entropy, median filtering
and simple time averaging in the window of “alive form”.
Working mode
–
the base window size can be assigned by the user or chosen
automatically;
Processing mode
–
two-dimensional filtering can be provided by layers and by cuts of the
source grid.
The filtering window sizes are chosen on the base of the analysis of three-dimensional autocorrelation function of
the source grid. The program creates the grid containing two signs: the local (remaining) and the regional
components of the source field.
5.3. Three-dimensional filtering...
In this section three-dimensional entropy and median filters are presented.
5.3.1. Three-dimension entropy filtering
Program is intended for three-dimensional filtering of the field.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Window width (in pickets)
–
the width of the filtration window (!!! If zero - defined automatically);
Window height (in profiles/layers) – the height of the filtration window (!!! If zero - defined automatically);
Window depth (in layers)
–
the depth of the filtration window (!!! If zero - defined automatically);
First incline of filtration window –
the slope of the filtration window in a plane of pickets and profiles;
Second incline of filtration window – the slope of the filtration window in a plane of pickets and layers;
Output grid
–
the result grid;
Filter type
–
there are two adaptive filters in this program: energy and median filtering.
The filtering window sizes are chosen on the base of the analysis of three-dimensional autocorrelation function of
the source grid. The program creates the grid containing two signs: the local (remaining) and the regional
components of the source field.
5.4. Decomposition of fields
In the given program automatic technology of the field decomposition by means of two-dimensional adaptive
filtering is marketed. The source information for this procedure is just the number to the source grid and the filter
type.
Input parameters of the module:
Input grid
–
Sign of input grid
–
Output grid
–
Copyright © 2003 MSGPU
the source grid;
the processed sign of the source grid;
the result grid;
Coscad 3Dt
Filter type
56
–
decomposition methods are based on different filters: energy, entropy,
median filtering and simple time averaging.
The program creates the grid containing several signs depending on presence of anomalies with different energy in
the source field. The first grid’s sign contains the most power-hungry forming (trend), the second sign contains the
smaller energy component and etc. The last sign contains the field component with amplitude commensurable with
noise level.
!!! ATTENTION !!! The input grid must be two-dimensional consisting of one layer. The number of pickets
and profiles must be more than 10!
5.5. Compensating filtering
The program is intended for solving of the problem of the exception from one field of correlation effect of the other
one.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Second input grid
–
the second source grid;
Sign of second input grid
–
the processed sign of the second source grid;
Base window width (in pickets) –
the width of the base filtration window;
Base window height (in profiles/layers) – the height of the base filtration window;
Frequency factor of the filter
–
the factor taking values from 0.5 up to 1.5;
If this parameter equals to 1, that spectrum of the analyzed frequencies is
defined just by size of base window and correlation score. Increasing of
this parameter allows to enlarge correlation coefficient.
Output grid
–
the result grid.
The program forms the grid including five signs. The first sign keeps the component of the first field uncorrelated
with the second one. The second sign (on the contrary) contains correlated component of the first field. The radius
of correlation is kept in the third sign. The fourth and fifth sign contains offsets by pickets and profiles of the
maximum two-dimensional cross-correlation functions accordingly.
5.6. Calculation of instantaneous power
This program estimates the instantaneous power in the slithering window along seismic traces (along the same
name pickets) i.e. across profile’s spreads.
Input parameters of the module:
Input grid
Sign of input grid
Second input grid
Sign of second input grid
Base width (in profiles)
Output grid
Copyright © 2003 MSGPU
–
–
–
–
–
–
the source grid;
the processed sign of the source grid;
the second source grid;
the processed sign of the second source grid;
the window width in profiles for the instantaneous power estimation;
the result grid.
Coscad 3Dt
57
6. DETECTION
Procedures of the program block "DETECTION" are intended for separation of weak geophisical anomalies of
linear and free form. In the terms of exploration geophysics the weak anomaly is accepted to consider as the signal
which is commensurable by amplitude with noise level or below of this level and its reliable visual finding is
practically impossible.
Amongst methods of the detection of weak anomaly algorithms built on theories of the statistical decisions taking
and checking of the statistical hypotheses are broadly used. Along with well-known algorithms of the detection: the
method of the inverse probabilities and self-tuning filtering, in this block enters adaptive algorithms taking into
account changing of parameters of anomalies and of the noise characteristics on area and in space. Except this, the
block contains the program realization of multivariate analogues of the inverse probability method and self-tuning
filtering increasing the separation possibility of weak geophisical signals onto several geophisical signs owing to
the accumulation of energy signal-to-noise merit.
Functional filling of the block "DETECTION" is effective for solving of the following geological problems:
 mapping of weakly shown tectonics in geophisical fields;
 separations of low-amperage lifting typical for hydrocarbon deposits;
 detection of weak anomalies conditioned by small-sizes geophisical objects located on small depth
(archeological and engineering geophysics).
6.1. Detection of weak linear anomalies...
Programs of detection of weak linear anomalies are enclosed in this block.
6.1.1. Method of interprofiles correlation
Input parameters of the module:
Input grid
Sign of input grid
Base width (in pickets)
Base height (in profiles)
Maximum slope of summation
–
–
–
–
–
the source grid;
the processed sign of the source grid;
the number of pickets in the base window;
the number of profiles in the base window;
the maximum offset within the extremum of CCF(m) is searching. This
parameter adjusts the possible offset at summation between profiles;
Incline of summation base
–
the slope of the summation base if it is assigned as constant;
Algorithm of base window choosing – the window is assigned or chosen automatically;
Algorithm of summation base choosing – the summation is assigned or chosen automatically;
Algorithm of window incline choosing – the slope is assigned or chosen automatically;
Output grid
–
the result grid.
Working mode
–
the base window size can be assigned by the user or chosen
automatically;
Processing mode
–
summation can be provided by layers and by cuts of the source grid.
The program realizes the method of interprofiles correlation in the slithering base window and is intended for
revealing and tracing (from profile to profile) of weak linear anomalies with amplitude commensurable with noise
level. This algorithm is concluded in a simple summation of field’s values by several neighboring profiles in the
slithering window with offset from profile to profile and normalization of the gotten sum on the number of the
profiles. The offset between profiles is chosen on values of the argument corresponding to the maximum of crosscorrelation function between them. The number of profiles of the summation is defined by the base window height,
which usually is taken to be equal from 5 to 31 profiles (or layers) depending on nature of the field and problems to
be solved. When the seismic information is processing the base parameters are usually assigned and chosen in
accordance with nature of the seismic cut.
Such directed summation of field values allows to emphasize existing and correlated from profile to profile weak
anomalies and reduce the influence of uncorrelated noise. The program creates the grid containing two signs - the
Copyright © 2003 MSGPU
Coscad 3Dt
58
result of the directed summation of the field and correlation direction (from profile to profile) in each point of the
source field.
6.1.2. Method of self-tuning filtering
This program realizes the so-called algorithm to self-tuning filtering, which is based on the calculation of the
Hotteling statistics that estimates the energy signal-to-noise merit.
Input parameters of the module:
Input grid
Sign of input grid
Filtering window width
Filtering window height
–
–
–
–
the source grid;
the processed sign of the source grid;
the width of the filtration window in pickets;
the height of the filtration window in profiles or layers;
The width is assigned much less than the height. Approximately in ratio
1:7 or 1:15.
Incline of filtering window
–
the slope of the filtering window;
Output grid
–
the result grid.
Method
–
it is possible to centralize data in the window (i.e. reduction to zero
average);
Working mode
–
the base window size can be assigned by the user or chosen
automatically;
Processing mode
–
filtration can be provided by layers and by cuts of the source grid;
Modification of the algorithm
–
the calculation of the maximum statistics is provided on different
directions or by total statistics. The last event is efficient when detecting
star-shaped structures at the searching of kimberlite pipes.
The Hotteling statistic is calculated in the slithering two-dimensional window of the fixed size with different
slopping. The slopping of the slithering window is chosen automatically or is assigned by the user as constant.
The assignment of the fixed slopping window is used when it is necessary to select anomalies with determined
direction. The window width must be much less than the window height (optimum is ratio 3х31). The algorithm is
efficient just for raw data in which the visual finding of anomalies does not possible. Otherwise, the usage of this
algorithm can be inefficient.
The program creates the grid in the database of the complex containing two signs: Hotteling statistic and correlation
direction of the field.
The statistic is interpreted in the following way: in points where its value is close to zero (less than 1.5) anomalies
are absent. The more the statistic’s values the more the probability of the presence of anomalies.
The correlation direction is possible to use at solving of the geological zoning problem using classification
programs.
6.1.3. Detection of multi-signs linear anomalies
This program is intended for revealing and tracing of weak complex linear anomalies commensurable by amplitude
with the noise level in conditions when the information about the form of the anomalies is absent.
Input parameters of the module:
Input grid
–
Amount of anomaly signs
–
Filtering window width
–
Filtering window height
Incline of filtering window
Output grid
Method
–
–
–
–
Copyright © 2003 MSGPU
the source grid;
the amount of signs in the multi-signs anomalies (concretely sign’s
numbers will be required in the dialogue after starting of the program);
the width of the filtration window in pickets. It is assigned much less than
window height. Usually it is 3-7 points;
the height of the filtration window in profiles or layers;
the slope of the filtering window;
the result grid;
it is possible to centralize data in the window (i.e. reduction to zero
average);
Coscad 3Dt
59
Slope mode
–
the slope can be assigned by the user or chosen automatically;
Processing mode
–
filtration can be provided by layers and by cuts of the source grid.
This program realizes the so-called multivariate analogue of self-tuning filtering, which is based on the calculation
of the spur of covariance matrix statistic in the slithering two-dimensional window with different slopping. The
Slope of the slithering window is chosen automatically in each point of the field or is assigned by the user as
constant for all points of the field. The assignment of the fixed slopping window is used when it is necessary to
select anomalies with determined direction. The window size is usually do not exceed seven pickets and nine
profiles. The algorithm is efficient just for raw data in which the visual finding of anomalies does not possible.
Otherwise, the usage of this algorithm can be inefficient.
The program creates the grid containing two signs: spur of covariance matrix statistic and correlation direction of
the field. The statistic is interpreted in the following way: in points where its value is close to zero (less than 1.5)
anomalies are absent. The more the statistic’s values the more the probability of the presence of anomalies.
The correlation direction is possible to use at solving of the geological zoning problem using classification
programs.
In contrast to univariate analogue of the method of self-tuning filtering the given algorithm possesses greater
reliability of the correct detection of anomalies. So, in the processing by two signs the reliability of the detection
increases not less than in four times, in the processing by three signs - increases not less than in nine times and etc.
Analyzed signs must be united in the one grid using the module "Uniting of grids" from the section "SERVICE" of
this program complex.
6.2. Detection of weak free form anomalies...
In the given block the program on detection of weak free form anomalies are enclosed.
6.2.1. Method of inverse probability
This program realizes the method of inverse probability and is used for detecting of weak one-signs or complex
anomalies in conditions when there is information on their form (the model anomaly). As the model anomaly most
often is taken the square-wave fragment to any grid from the database of the complex where the presence of the
anomaly is realistically known.
Input parameters of the module:
Input grid
–
Second input grid
–
the source grid;
the source of the model anomaly. The grid containing the anomaly
fragment (may comply with the source grid);
Left picket of anomaly
–
the left border of the anomaly by pickets;
Right picket of anomaly
–
the right border of the anomaly by pickets;
Upper profile (layer) of anomaly –
the upper border of the anomaly by profiles (layers);
Bottom profile (layer) of anomaly –
the bottom border of the anomaly by profiles (layers);
Amount of anomaly signs
–
the amount of signs in the multi-signs anomalies (concretely sign’s
numbers will be required in the dialogue after starting of the program);
Output grid
–
the result grid;
Method
–
it is possible to centralize data in the window (i.e. reduction to zero
average);
Processing mode
–
filtration can be provided by layers and by cuts of the source grid.
For each point of the field the spur of covariance matrix statistic F is computed, which is a multivariate analogue of
the Fisher’s statistic.
Values of the F-statistic is interpreted in the following way: If its values is located in the interval 0<=F<=0.5 that
the model anomaly in the point is absent. If the value of this statistic falls into interval 0.5<F<=1.5 that the anomaly
is presented. And finally if the value of the statistic is 1.5<F<1000 - the model anomaly is present and its form
complies with model, but the amplitude exceeds the amplitude of model anomaly.
The usage of this module is effective just when visual finding of anomalies are practically impossible (for example,
in processing of uncorrelated field both on pickets and on profiles). Most often, the last local field component has
characteristic like this.
Copyright © 2003 MSGPU
Coscad 3Dt
60
The program creates the grid containing Fisher’s statistic F.
Analyzed signs must be united in the one grid using the module "Uniting of grids" from the section "SERVICE" of
this program complex.
6.2.2. Automatic variant of the method of inverse probability
This program realizes the modified algorithm of the method of inverse probability allowing to select four types
(positive, negative and two alternating-sign types - plus-minus and minus-plus) of weak anomalies commensurable
with noise level in the automatic mode.
Input parameters of the module:
Input grid
–
the source grid;
Sign of input grid
–
the processed sign of the source grid;
Output grid
–
the result grid;
Method
–
it is possible to centralize data in the window (reduction to zero average);
Processing mode
–
filtration can be provided by layers and by cuts of the source grid.
The program creates a four-signs grid. The first sign contains the F-statistic for checking the hypothesis about the
positive anomaly presence, the second sign - about the negative anomaly presence, the third – about alternatingsign “plus-minus” anomalies, the fourth - alternating-sign “minus-plus” anomalies.
Values of the F-statistic is interpreted in the following way: If its values is located in the interval 0<=F<=0.5 that
the model anomaly in the point is absent. If the value of this statistic falls into interval 0.5<F<=1.5 that the anomaly
is presented. And finally if the value of the statistic is 1.5<F<1000 - the model anomaly is present and its form
complies with model, but the amplitude exceeds the amplitude of model anomaly.
In contrast to the classical variant of the method of inverse probability this algorithm does not require presence to
model anomaly.
Analyzed signs must be united in the one grid using the module "Uniting of grids" from the section "SERVICE" of
this program complex.
Copyright © 2003 MSGPU
Coscad 3Dt
61
7. COMPLEX
The functional block "COMPLEX" unites the procedures oriented on the processing of multi-signs geologygeophisical information by methods of classifications, patterns recognition and component-analysis. As input
information for programs of this group the values of different geology-geophisical signs and their derived gotten by
means of programs from the other sections of the complex can be used. Most algorithms included in this block are
based on checking of the multivariate statistical hypothesis that allows to use information on structure of multisigns relationship between different geophisical fields.
Procedures of this block are directed on efficient solving of the following problems:
 partition of the analyzed area onto uniform on several signs areas (geological zoning);
 recognition of multi-signs geophisical anomalies with the presence of a priori information about the form of
model multi-signs anomaly;
 analysis of multi-signs information by means of classical component-analysis;
 optimum scaling (the digitization) of nominal geology-geophisical signs;
 estimations of self-descriptiveness of separate signs and their different combination.
Classification algorithms being included in this block are built in principles of self-training, account of correlation
relationships of whole sign’s space and possibility of the correct work in conditions of the absence of a priori
information about initial classes center’s positions and final amount of uniform areas.
For recognition algorithms the source information is just a form to anomalies, but sometimes when the model is a
fragment of under investigation area, only a sidebar of multi-signs anomaly. Estimation of a multivariate noise
parameter and its influence is excluded right in the process of the recognition. The parameter characterizing the
model object for algorithms of the recognition are two-dimensional surfaces given in discrete points of the
observations of the square-wave grid. Each surface reflects the form of the concrete physical field above the model
object. Signs (in general) can be expressed by different geophisical fields, their derived, observations of the field on
different levels, digitized geological, petrophysical and geochemical information.
7.1. Classification...
In the given section different classification algorithms allowing splitting the under investigation territories into
areas uniform by complex of signs are enclosed.
7.1.1. Method of the common distance
This program is intended for classification of multi-signs data by the method of the common distance, which is
intended for partition of the under investigation territory into areas uniform by complex of signs. The final amount
of classes (uniform areas) is chosen automatically.
Input parameters of the module:
Input grid
–
Amount of signs
–
the source grid;
the amount of signs (concretely sign’s numbers will be required in the
dialogue after starting of the program);
Output grid
–
the result grid;
Metrics
–
it is possible to classify in Euclidean metric L2 and in metric L1.
As a source sign you can use not only different geology-geophisical observations, but also secondary data of one
sign. For example, it can be gradients of the field and statistical moments and etc. Analyzed signs must be united in
the one grid using the module "Uniting of grids" from the section "SERVICE" of this program complex.
The program creates the grid containing the results of the classification. Each point of the grid is characterized by
number of the class, to which it belongs.
Besides a file is creating in the active directory “*.CLS” (the asterisk is the number to the resulting grid. For
example 23.CLS - for 23rd grid) in which information about average value, standard deviation and the standard of
separate signs in each class and quality of the classification is written. The closer the last parameter to 1 the better
quality of the classification.
Copyright © 2003 MSGPU
Coscad 3Dt
62
This program is efficient in solving of problems of geological mapping and zoning by complex of parameters.
7.1.2. Method of the dynamic thickenings (К-mean-value method)
This program is intended for classification of multi-signs data by method of the dynamic thickenings, which is
intended for partition of the under investigation territory into areas uniform by complex of signs. User assigns
information about a final classes amount (uniform areas).
Input parameters of the module:
Input grid
–
Amount of signs
–
the source grid;
the amount of signs (concretely sign’s numbers will be required in the
dialogue after starting of the program);
Target amount of classes
–
the final number of uniform areas;
Output grid
–
the result grid;
Metrics
–
it is possible to classify in Euclidean metric L2 and in metric L1.
As a source sign you can use not only different geology-geophisical observations, but also secondary data of one
sign. For example, it can be gradients of the field and statistical moments and etc. Analyzed signs must be united in
the one grid using the module "Uniting of grids" from the section "SERVICE" of this program complex.
The program creates the grid containing the results of the classification. Each point of the grid is characterized by
number of the class, to which it belongs.
Besides a file is creating in the active directory “*.CLS” (the asterisk is the number to the resulting grid. For
example 23.CLS - for 23rd grid) in which information about average value, standard deviation and the standard of
separate signs in each class and quality of the classification is written. The closer the last parameter to 1 the better
quality of the classification.
This program is efficient in solving of problems of geological mapping and zoning by complex of parameters.
7.1.3. Classification by A.V. Petrov
The program realizes a division of the multivariate normal mixtures algorithm, which is intended for partition the
territory into areas uniform by complex of signs, with automatic determination of the final classes amount (also it is
possible to assign the final classes amount by user).
Input parameters of the module:
Input grid
–
Amount of signs
–
the source grid;
the amount of signs (concretely sign’s numbers will be required in the
dialogue after starting of the program);
Target amount of classes
–
the final number of uniform areas;
Output grid
–
the result grid;
Working mode
–
it is possible to switch on the mode of automatic determination of the
final amount of classes.
As a source sign you can use not only different geology-geophisical observations, but also secondary data of one
sign. For example, it can be gradients of the field and statistical moments and etc. Analyzed signs must be united in
the one grid using the module "Uniting of grids" from the section "SERVICE" of this program complex.
The program creates the grid containing the results of the classification. Each point of the grid is characterized by
number of the class, to which it belongs.
Besides a file is creating in the active directory “*.CLS” (the asterisk is the number to the resulting grid. For
example 23.CLS - for 23rd grid) in which information about average value, standard deviation and the standard of
separate signs in each class and quality of the classification is written. The closer the last parameter to 1 the better
quality of the classification.
This program is efficient in solving of problems of geological mapping and zoning.
Copyright © 2003 MSGPU
Coscad 3Dt
63
7.1.4. Sign classification
!!! (In this section there is one tautology: We use the word “sign” in two meanings. The first
is a common of this manual meaning “sign” – a one characteristic measured in each point of
the grid. The second one is a math sign meaning like plus or minus. – translator’s mark) !!!
This program realizes the algorithm allowing to split the processed grid on areas having alike signs of the field by
complex of signs. For example if two signs are processing then the program will choose areas where both signs
positive (+ +), the first - positive and the second is negative (+ -), the first - negative and the second is positive (- +)
and finally, both signs are negative (- -) (it is naturally that all analyzed signs are centered beforehand). If an
amount of analyzed signs is n than amount of areas is defined by expression 2n.
Input parameters of the module:
Input grid
–
Amount of signs
–
the source grid;
the amount of signs. !not more than 8! (concretely sign’s numbers will
be required in the dialogue after starting of the program);
Output grid
–
the result grid;
Besides a file is creating in the active directory “*.CLS” (the asterisk is the number to the resulting grid. For
example 23.CLS - for 23rd grid) in which information about average value, standard deviation and the standard of
separate signs in each class and quality of the classification is written.
In spite of simplicity of the applicable algorithm, its efficacy in solving of problems of geological mapping and
zoning is obvious. In case of big size of information and amount of analyzed signs a such partition is practically
impossible to conduct without a computer.
7.2. Detection of multi-signs anomalies
This program is intended for detection of multi-signs geophisical anomalies.
Input parameters of the module:
Input grid
–
Second input grid
–
the source grid;
the source of the model anomaly. The grid containing the anomaly
fragment (may comply with the source grid);
Left picket of anomaly
–
the left border of the anomaly by pickets;
Right picket of anomaly
–
the right border of the anomaly by pickets;
Upper profile (layer) of anomaly –
the upper border of the anomaly by profiles (layers);
Bottom profile (layer) of anomaly –
the bottom border of the anomaly by profiles (layers);
Amount of anomaly signs
–
the amount of signs in the multi-signs anomalies (concretely sign’s
numbers will be required in the dialogue after starting of the program);
Output grid
–
the result grid;
Method
–
it is possible to centralize data in the window (i.e. reduction to zero
average);
Working mode
–
data in the window can be smoothed for the removing of the highfrequency noise reasons;
Processing mode
–
filtration can be provided by layers and by cuts of the source grid.
For realization of the algorithm information about model anomaly is necessary. This information can be taken from
the source or any other grid. As a model anomaly most often is taken the square-wave fragment of any grid from
the database of the complex including the source one. For example, square-wave fragment of the field around an
ore bore hole can be used like model anomaly.
The algorithm is based on calculation of the spur of covariance matrix statistic in the slithering window of the fixed
sizes with different slopping. The calculation of the statistic under different slopping of the window allows to
localize areas where fields are "similar" with model, but their orientation differs from model.
The program forms the grid with an amount of signs equals to an amount of directions assigned by the user. Each
sign contains a spur of covariance matrix statistic corresponding to each direction.
Copyright © 2003 MSGPU
Coscad 3Dt
64
The statistic’s values are interpreted as follows: The more similar fields in the vicinity of the point onto the model
the less the value of the statistic. Usually, the decision threshold about the presence of the model anomaly is not
more than three. It is considered that in the vicinity of the point the probability of the model anomaly presence is
high if the calculated statistic of this vicinity lies within the range from 0 up to 2.5.
The algorithm is efficient in solving of problems of the forecasting, geological mapping and zoning.
7.3. Components data analysis
This program is intended for carrying out of the classical components analysis of multi-signs geophisical
information by means of covariance matrix calculation, searching for eigenvalues and eigenvectors and the further
folding of multi-signs information by eigenvectors.
Input parameters of the module:
Input grid
–
Amount of signs
–
the source grid;
the amount of signs in the multi-signs grid (concretely sign’s numbers
will be required in the dialogue after starting of the program);
Output grid
–
the result grid;
Amount of signs in the resulting grid complies with amount of analyzed signs. The first sign (with the bigest
dispersion) presents the folding of analyzed signs with the covariance matrix corresponding to the maximum
eigenvalue, the second sign (with a little bit smaller dispersion) - presents the folding with eigenvectors
corresponding to following in size eigenvalue and etc.
Copyright © 2003 MSGPU